Global Guide to Autonomous Vehicles
Discover the guides in 8 jurisdictions with our interactive color coded map. Click on a jurisdiction to access its dedicated page.
Explore
Welcome to the Dentons Global Autonomous Vehicle Guide, an interactive resource designed to help you explore the rapidly evolving autonomous vehicle landscape around the world. Browse country-by-country guides to better understand the laws, regulations, and policy developments shaping the future of autonomous mobility.
Welcome
Canada
India
Germany
United States of America
Australia
United Kingdom
China
South Korea
View the full guide here!
2026 US Autonomous Vehicles Guide: Navigating the Legal and Regulatory Landscape
Authors
Eric Tanenblatt Head, Global Autonomous Vehicles and Principal, Washington DC and Atlanta eric.tanenblatt@dentons.com
Brett Dorman Partner, Los Angeles brett.dorman@dentons.com
Walker Boothe Associate Managing Director, Atlanta walker.boothe@dentons.com
Jordan Cooper Managing Director, Washington DC jordan.cooper@dentons.com
Todd Daubert Partner, Washington DC todd.daubert@dentons.com
Jeff Denham Senior Policy Director, Washington DC jeff.denham@dentons.com
View guide
© 2026 Dentons. All rights reserved. Attorney Advertising. Dentons is a global legal practice providing client services worldwide through its member firms and affiliates. This website and its publications are not designed to provide legal or other advice and you should not take, or refrain from taking, action based on its content.
Ben Allen Partner, Sydney ben.allen@dentons.com
Thuy Hoang Business Development Advisor, Sydney thuy.hoang@dentons.com
Hugh Cranendonk Senior Associate, Sydney hugh.cranendonk@dentons.com
Amer Pasalic Partner, Toronto amer.pasalic@dentons.com
Martin Abadi Partner, Toronto martin.abadi@dentons.com
Henry Chen Senior Partner, Shanghai henry.chen@dentons.cn
Link Lin Partner, Shanghai link.lin@dentons.cn
Simon Zhu Lawyer, Shanghai simon.zhu@dentons.cn
Darian Cai Lawyer, Shanghai darian.cai@dentons.cn
Kartikeya Singh Principal Associate, New Delhi kartikeya.singh@dentonslinklegal.com
Anuj Trivedi Partner, New Delhi anuj.trivedi@dentonslinklegal.com
Shashwat Sinha Associate, New Delhi shashwat.sinha@dentonslinklegal.com
Dr. Michael Malterer Partner, Munich michael.malterer@dentons.com
Kurt Gerstner Senior Attorney Admitted to New York, Massachusetts and DC Bars, Seoul kurt.gerstner@dentons.com
Matthew Gilhooly Senior Associate, Glasgow matthew.gilhooly@dentons.com
Australia is navigating a critical transitional period in its autonomous vehicle journey. The regulatory framework that will govern the safe deployment of automated vehicles is still being finalized, but progress is accelerating at both the Commonwealth and state level. The overarching policy direction is set by the 2024-27 Automated Vehicle Action Plan (CAV Action Plan), which establishes a structured, nationally coordinated program of reform designed to bring Australia into step with leading international AV jurisdictions. The Action Plan implements the National Road Transport Technology Strategy (NTTS Strategy) to prepare Australia for the deployment of Connected and Automated Vehicles (CAVs), including Cooperative Intelligent Transport Systems (C-ITS). Its key priorities are: Completing the end-to-end AV regulatory framework; advancing nationally consistent C-ITS deployment; and addressing cross-cutting issues including data, workforce, infrastructure and accessibility.[1] Readers of the 2023 edition of the Global AV Guide will recall that the Automated Vehicle Safety Law (AVSL) was expected to commence by 2026.[2] That timeline was not met. Transport Ministers agreed on November 21, 2025, to allow conditional deployment of automated vehicles from 2027 in selected locations, with full national readiness to follow.[3] All states and territories have trialed CAVs, with Queensland undertaking the largest on-road CAV and C-ITS trial to date. The Ipswich Connected Vehicle Pilot found that, at 100% C-ITS penetration, the use cases investigated could reduce fatal and serious injury crashes by 13 to 20 percent. By 2050, CAVs are estimated to reduce crash costs by approximately $15 billion, increasing to $152 billion by 2070.[4]
What are the latest autonomous vehicle developments in 2025 and early 2026?
1.
Guide landing page
[1] Department of Infrastructure, Transport, Regional Development, Communications and the Arts (Cth), 2024–27 National Connected and Automated Vehicle (CAV) Action Plan (Policy Document, 2024); Department of Infrastructure, Transport, Regional Development, Communications and the Arts (Cth), National Road Transport Technology Strategy (Strategy, 2023). [2] Dentons, Dentons Global Guide to Autonomous Vehicles 2023 (Web Document, April 2023) Australia's primary roadblock to mainstream AV deployment is the challenge of harmonizing road traffic legislation across its eight states and territories into a single, nationally consistent regulatory environment. Unlike jurisdictions with unified national traffic law, road rules and motor vehicle legislation in Australia are principally state and territory matters. This means that any effective AV legal framework requires coordinated legislative reform across multiple jurisdictions before a vehicle can operate seamlessly on roads nationwide. The CAV Action Plan has been designed specifically to address this challenge, but it remains the central structural constraint on large-scale cross-jurisdictional AV deployment in Australia. Vehicles are classified into six different levels of automation as developed by the Society of Automotive Engineers (SAE). The six levels are as follows: Is your country experiencing any recent roadblocks regarding autonomous vehicle developments? 2. [5] Ben Allen, 'Safety Meets Privacy: What Australia's Automated Vehicle Laws Mean for Your Business' (Web Page, Dentons, 23 April 2026). [6] Ibid. [7] Ibid. [8] National Road Transport Technology Strategy (n 1). [9] Ben Allen (n 5). Australia's existing privacy and security legal framework already imposes significant obligations on AV operators and developers. The intersection of safety regulation and privacy law is emerging as one of the most practically significant legal challenges for the sector, and compliance complexity will only grow as vehicles become more connected and more data-intensive. Safety compliance and privacy compliance can no longer be treated as separate workstreams. The data that makes automated vehicles safe is often the same data that triggers privacy obligations.[5] Under the Privacy Act 1988 (Cth), the threshold for personal information is easily met in an AV context. Categories commonly generated include GPS data, driver behavior data, biometric data, internal and external camera recordings, and data from connected apps. Some may constitute sensitive information, attracting higher legal standards.[6] The AVSL will cover vehicles supported or operated remotely. Without a clear data map, compliance is extremely difficult; organizations need to understand how data moves from collection to sharing with service providers or regulators before meaningful compliance is achievable.[7] Action 2.3 of the CAV Action Plan will develop a national system to manage C-ITS message security, with a business case expected by 2026. Existing security obligations may already apply to operators, including under the Telecommunications (Interception and Access) Act 1979 (Cth), the Telecommunications Act 1997 (Cth) and the Security of Critical Infrastructure Act 2018 (Cth).[8] The privacy regulator has publicly identified connected vehicles as an area of concern.[9] Are there any recent updates in the cybersecurity/privacy space related to autonomous vehicles? 3. [10] 2024–27 National Connected and Automated Vehicle (CAV) Action Plan (n 1); National Road Transport Technology Strategy (n 1). [11] Shuo Ding, Aidin Bervan and Erik van Vulpen, Automated Mobility for Elderly and People with Disability: How 5G Can Help (Project Report, iMOVE CRC and La Trobe University Centre for Technology Infusion, 14 December 2022). [12] 2024–27 National Connected and Automated Vehicle (CAV) Action Plan (n 1). Australia's telecommunications infrastructure is increasingly being recognized as a foundational enabler of CAV deployment. Australian researchers and policymakers are actively working to ensure that the country's spectrum arrangements and connectivity standards keep pace with international developments, and on-road trials are beginning to generate a domestic evidence base for policy decisions. Our 2023 submission covered the Australian Communications and Media Authority’s Radiocommunications (Intelligent Transport Systems) Class License 2017, under which 70 MHz of spectrum in the 5.9 GHz band was made available for C-ITS. The 2024–27 CAV Action Plan has since introduced specific actions to maintain alignment with international markets. Action 2.4 monitors radiofrequency spectrum arrangements for C-ITS, and Action 2.6 requires ADRs to be updated as international connectivity standards develop. Existing Australian telecommunications infrastructure, including 4G, 5G and Low Earth Orbit satellite networks, could play a role in connecting CAVs. iMOVE Australia microsimulation research suggests C-ITS could reduce peak arterial corridor congestion by 11 percent and improve peak-hour travel speeds in Melbourne's CBD by 10 percent.[10] The practical performance of these networks in an Australian AV context has also been the subject of dedicated on-road research as previously discussed. Researchers at La Trobe University's Centre for Technology Infusion, funded by iMOVE CRC, trialed 4G and 5G communications on an autonomous vehicle at Sydney Olympic Park and Bundoora. The trials demonstrated that Roadside Units (RSUs) communicating via 4G and 5G can significantly improve obstacle detection, identifying hazards from at least 35 meters compared to only 11 meters using on-board LiDAR, with 5G remote monitoring achieving end-to-end latency of approximately 150ms. The trials provide an Australian evidence base for the national C-ITS roadmap that Action 2.1 is now developing.[11] Action 2.1 requires the development and publication of a national C-ITS roadmap enabling a staged, nationally consistent approach. Lastly, Action 2.5 sets out a mechanism to identify impediments to implementing the eCall in-vehicle automated crash notification system in Australia.[12] Are there any recent updates in the 5G space related to autonomous vehicles? 4. [13] Ben Allen (n 7) As noted above, the AVSL was originally expected to be in force by 2026, but that timeline was not achieved. The CAV Action Plan represents Australia's current legislative roadmap, designed to complete the end-to-end AV regulatory framework and enable safe commercial deployment by 2027. It sets out a comprehensive program of legislative and policy work at Commonwealth, state and territory level, and is the most significant structural step forward in Australia's AV regulatory journey to date. Once in effect, the AVSL will place responsibility for AV safety on a corporation, the Automated Driving System Entity (ADSE), rather than a human driver. The AVSL will be supported by complementary Commonwealth and state legislative changes, forming an end-to-end framework covering first supply, ADSE capability, on-road safety duties and user responsibilities ADSE status is not determined solely by who built the technology; the key question is whether an organization has sufficient influence over safety-critical decisions. Many businesses, including software providers, fleet operators and remote monitoring providers, will need to assess their position carefully in light of this framework.[13] Have any new laws/regulations been passed within the last year that have affected the deployment of autonomous vehicles? 5. Key CAV Action Plan actions timed for 2026 include: Action 2.3(b): the development of a business case for a national C-ITS security credential management system, which will provide a coordinated national approach to authenticating and securing communications between connected vehicles and roadside infrastructure, a foundational requirement before large-scale C-ITS deployment can proceed; Action 2.4: continued monitoring of radio frequency spectrum arrangements for C-ITS to ensure they remain fit for purpose and appropriately aligned with key international markets, covering both short-range and long-range connectivity technologies; Action 3.6: investigation of appropriate in-vehicle alert standards, including visual, auditory and haptic (touch-based) systems that communicate the operational status of automated driving systems to occupants and guide appropriate human responses, which is particularly important for ensuring safety at the boundary between human-driven and automated modes; Action 3.9: investigation of a national coordination role for Mobility as a Service (MaaS) platform, considering how automated vehicles can be integrated across transport modes and providers to improve accessibility, efficiency and service delivery, particularly for users who do not or cannot drive; and Action 3.10: identification of strategies for reducing greenhouse gas emissions through optimal CAV deployment, recognizing that the efficiency gains from connected and automated driving could contribute meaningfully to Australia's broader transport emissions reduction objectives. Work has also commenced on updated state and territory road traffic laws to regulate human users of AVs. Action 1.8 captures the National Transport Commission's regular readiness assessments, timed for 2025 and 2027. What advancements in autonomous vehicles should we expect from your country in 2026? 6. Australia's geography and industrial profile make it one of the most compelling environments in the world for commercial AV adoption beyond the passenger vehicle context. The dispersed population centres, significant physical separation between major metropolitan areas, and the scale and remoteness of Australia's mining operations, agricultural landholdings and freight corridors create both a strong economic case and a practical necessity for automation. Australia's mining industry, concentrated in remote regions of Western Australia and Queensland, has long operated at the frontier of heavy vehicle automation. Its agricultural sector faces comparable pressures: chronic labor shortages, vast broad acre properties spanning hundreds of thousands of hectares, and the imperative to maximize agricultural output with fewer resources have similarly driven substantial investment in autonomous and semi-autonomous machinery. The logistics and freight industries face a parallel challenge. Servicing widely separated urban markets across the country demands supply chain solutions that human-operated transport alone struggles to deliver at scale. Commercial AV deployment in Australia is therefore not simply a question of technological ambition, but increasingly a matter of industrial necessity. The following examples illustrate how that necessity is already translating into real-world deployment across four key sectors: Road and Freight Trucking A multinational vehicle manufacturer and an Australian motorway operator commenced a trial in August 2025 of C-ITS technology on Australian motorways. The motorway operator has also partnered with an international autonomous driving software company, building on a 2022 self-driving truck trial on major Melbourne motorways, to explore how Level 4 autonomous driving technology combined with smart road infrastructure could make heavy freight trucking safer and more efficient. Mining There are already automated trucks operating on mine sites in north-west Western Australia. A multinational truck manufacturer is preparing to launch what is described as the world's first fleet of autonomous in-pit mining trucks at an Australian manganese mine, where autonomous rigid tippers will transport ore across the site. Farming Australia's agricultural sector has emerged as a significant arena for autonomous vehicle deployment, driven by acute labor shortages and the operational demands of vast broad acre properties. GPS-guided autonomous tractors, self-driving seeders and robotic harvesting machinery operate increasingly without human intervention on large-scale grain and cotton farms, with major equipment manufacturers having introduced autonomous and semi-autonomous combines to the Australian market and the Government's On Farm Connectivity Program, providing dedicated funding support for the connectivity infrastructure on which these systems depend. Warehousing and Industrial Logistics Autonomous vehicle technology is increasingly being adopted in controlled, off-road environments such as warehouses, factory floors and mine sites. Australian operators are actively deploying purpose-built autonomous vehicles in industrial and logistics settings. Because these vehicles operate off public roadways, they fall outside the scope of the AVSL framework and are instead governed by existing workplace health and safety obligations. Aside from robotaxis, do you see any other developments in the commercial space regarding AVs, including farming, commercial trucking, food delivery, etc.? 7. [14] Department of Industry, Science and Resources (Cth), Safe and Responsible AI in Australia: Australian Government's Interim Response (Report, 2024). [15] Safe and Responsible AI in Australia: Australian Government's Interim Response (n 16); Ben Allen (n 7). [16] 2024–27 National Connected and Automated Vehicle (CAV) Action Plan (n 1). Australia has not yet enacted standalone AI-specific legislation. However, the interaction between AI governance and the AV regulatory framework is a live and developing area, with the government having classified AI-enabled vehicles as a higher-risk application. In the current absence of dedicated AI legislation, AV operators must currently navigate a combination of existing sector-specific laws and the AVSL framework. Further clarity is expected as Action 3.1 of the CAV Action Plan progresses. It is also important to note that the artificial intelligence underpinning autonomous vehicles is fundamentally different in character from the large language model AI that has attracted significant public and regulatory attention in recent years. AV artificial intelligence does not operate primarily by processing and generating natural language. Rather, it is a complex integration of neural networks, machine learning algorithms and real-time sensor data-processing models that must interpret dynamic physical environments, make safety-critical decisions within milliseconds and interact with unpredictable road conditions at speed. This technical distinction has direct regulatory implications. Legislative frameworks designed with generative AI in mind may not be well-suited to addressing the specific risk profile of AV systems, and any future Australian AI regulation will need to account for this difference if it is to engage meaningfully with the autonomous vehicle sector. AVs may be enabled by AI, and the government has classified this as a higher-risk application. AI-related concerns relevant to AV deployment are addressed through the existing cross-cutting legal framework and the holistic law review contemplated by Action 3.1.[14] Safety and privacy compliance can no longer be treated as separate workstreams. The ADSE is the organization legally responsible for the safety of an automated driving system throughout its operational life, and ADSE status hinges on whether an organization has sufficient influence over safety-critical decisions, making governance of the underlying AI system a central question under the AVSL framework.[15] Australia is participating in United Nations work to develop international AV standards, which will form the basis of ADRs. This is the primary mechanism through which international AI governance principles may influence the Australian AV regulatory framework in the absence of domestic AI-specific legislation. As of early 2026, mandatory AI-specific regulation had not been enacted. Whether any future AI regulation will intersect with the AVSL framework remains to be resolved under Action 3.1 of the CAV Action Plan.[16] Have any AI laws impacted deployment? 8. Australia's approach to AV regulation is, in substance, a deliberate "wait and see" posture. Rather than moving to enact prescriptive AV-specific legislation ahead of widespread deployment, the government has chosen to observe how the technology matures and comparable jurisdictions, most notably the United States, resolve the legal, technical and policy tensions that AV integration presents. The primary evidence for this characterization is the trajectory of the AVSL framework itself. Despite years of consultation and the publication of successive action plans, the framework has not been fully operationalized, and the Government's own 2026 implementation milestone has not been met. Australia's regulatory position is not, however, shaped by legislation alone. The pace and direction of AV integration in Australia is also being guided by domestic industrial needs, as evidenced by the rapid uptake of autonomous vehicle technology in the mining, agricultural and logistics sectors, and by shifting consumer trends, including growing familiarity with ADAS features and increasing public exposure to AV technology through internationally available consumer vehicles. This cautious incrementalism may reflect a pragmatic recognition that legislating ahead of the technology carries its own risks. It also means, however, that Australia enters this period of rapid commercial AV development without the regulatory certainty that both industry and the public will ultimately require. Conclusion SAE Level Name Description Example 0 No automation Features limited to providing warnings and momentary assistance. The human driver performs all aspects of the driving task. Lane departure warning 1 Driver assistance Features provide steering or brake/acceleration support to the driver. Adaptive cruise control 2 Partial driving automation Features provide a combination of steering and brake/acceleration support. The human driver remains responsible for all other aspects of the driving task. Lane keeping and adaptive cruise control simultaneously 3 Conditional driving automation The vehicle can drive itself under limited conditions, but a user must be receptive to requests to intervene when required. Traffic jam chauffeur 4 High-driving automation The vehicle can drive itself under limited conditions with no expectation that a user will need to intervene. Driverless taxi operating within a defined geographic area 5 Full driving The vehicle can drive itself under all conditions without any expectation that a user will need to intervene. Driverless taxi without operational limitations Many vehicles on Australian roads already have Advanced Driver Assistance Systems (ADAS) such as lane keeping assist, assisted breaking and adaptive cruise control, but these do not make a vehicle "automated" in the regulatory sense. The critical legal threshold is the transition from Level 2 to Level 3, at which point the automated driving system takes over and human responsibility begins to diminish. Current Australian Road Rules were developed for SAE Levels 0-2, and the Automated Vehicle Program is focused on updating them to accommodate Levels 3 and above. Achieving this requires not only Commonwealth legislative action, but also parallel amendments to road traffic laws in each state and territory, making the harmonization process both technically complex and politically challenging. The CAV Action Plan addresses this through: Action 1.4, requiring states and territories to update their road traffic laws and enforcement activities to support automated vehicle operations; Action 1.5, monitoring motor accident injury insurance scheme changes across jurisdictions to align with the AVSL; and Action 3.1, holistically assessing existing Commonwealth, state and territory laws and identifying legislative gaps that could impede nationally consistent deployment. Vehicle connectivity standards will be translated to Australian Design Rules (ADRs) as they are finalized internationally, with a focus on ensuring imported vehicles can use Australia's C-ITS systems. Back to top of page Back to top of page
Australia's primary roadblock to mainstream AV deployment is the challenge of harmonizing road traffic legislation across its eight states and territories into a single, nationally consistent regulatory environment. Unlike jurisdictions with unified national traffic law, road rules and motor vehicle legislation in Australia are principally state and territory matters. This means that any effective AV legal framework requires coordinated legislative reform across multiple jurisdictions before a vehicle can operate seamlessly on roads nationwide. The CAV Action Plan has been designed specifically to address this challenge, but it remains the central structural constraint on large-scale cross-jurisdictional AV deployment in Australia. Vehicles are classified into six different levels of automation as developed by the Society of Automotive Engineers (SAE). The six levels are as follows:
Is your country experiencing any recent roadblocks regarding autonomous vehicle developments?
2.
[5] Ben Allen, 'Safety Meets Privacy: What Australia's Automated Vehicle Laws Mean for Your Business' (Web Page, Dentons, 23 April 2026). [6] Ibid. [7] Ibid. [8] National Road Transport Technology Strategy (n 1). [9] Ben Allen (n 5).
Australia's existing privacy and security legal framework already imposes significant obligations on AV operators and developers. The intersection of safety regulation and privacy law is emerging as one of the most practically significant legal challenges for the sector, and compliance complexity will only grow as vehicles become more connected and more data-intensive. Safety compliance and privacy compliance can no longer be treated as separate workstreams. The data that makes automated vehicles safe is often the same data that triggers privacy obligations.[5] Under the Privacy Act 1988 (Cth), the threshold for personal information is easily met in an AV context. Categories commonly generated include GPS data, driver behavior data, biometric data, internal and external camera recordings, and data from connected apps. Some may constitute sensitive information, attracting higher legal standards.[6] The AVSL will cover vehicles supported or operated remotely. Without a clear data map, compliance is extremely difficult; organizations need to understand how data moves from collection to sharing with service providers or regulators before meaningful compliance is achievable.[7] Action 2.3 of the CAV Action Plan will develop a national system to manage C-ITS message security, with a business case expected by 2026. Existing security obligations may already apply to operators, including under the Telecommunications (Interception and Access) Act 1979 (Cth), the Telecommunications Act 1997 (Cth) and the Security of Critical Infrastructure Act 2018 (Cth).[8] The privacy regulator has publicly identified connected vehicles as an area of concern.[9]
Are there any recent updates in the cybersecurity/privacy space related to autonomous vehicles?
3.
[10] 2024–27 National Connected and Automated Vehicle (CAV) Action Plan (n 1); National Road Transport Technology Strategy (n 1). [11] Shuo Ding, Aidin Bervan and Erik van Vulpen, Automated Mobility for Elderly and People with Disability: How 5G Can Help (Project Report, iMOVE CRC and La Trobe University Centre for Technology Infusion, 14 December 2022). [12] 2024–27 National Connected and Automated Vehicle (CAV) Action Plan (n 1).
Australia's telecommunications infrastructure is increasingly being recognized as a foundational enabler of CAV deployment. Australian researchers and policymakers are actively working to ensure that the country's spectrum arrangements and connectivity standards keep pace with international developments, and on-road trials are beginning to generate a domestic evidence base for policy decisions. Our 2023 submission covered the Australian Communications and Media Authority’s Radiocommunications (Intelligent Transport Systems) Class License 2017, under which 70 MHz of spectrum in the 5.9 GHz band was made available for C-ITS. The 2024–27 CAV Action Plan has since introduced specific actions to maintain alignment with international markets. Action 2.4 monitors radiofrequency spectrum arrangements for C-ITS, and Action 2.6 requires ADRs to be updated as international connectivity standards develop. Existing Australian telecommunications infrastructure, including 4G, 5G and Low Earth Orbit satellite networks, could play a role in connecting CAVs. iMOVE Australia microsimulation research suggests C-ITS could reduce peak arterial corridor congestion by 11 percent and improve peak-hour travel speeds in Melbourne's CBD by 10 percent.[10] The practical performance of these networks in an Australian AV context has also been the subject of dedicated on-road research as previously discussed. Researchers at La Trobe University's Centre for Technology Infusion, funded by iMOVE CRC, trialed 4G and 5G communications on an autonomous vehicle at Sydney Olympic Park and Bundoora. The trials demonstrated that Roadside Units (RSUs) communicating via 4G and 5G can significantly improve obstacle detection, identifying hazards from at least 35 meters compared to only 11 meters using on-board LiDAR, with 5G remote monitoring achieving end-to-end latency of approximately 150ms. The trials provide an Australian evidence base for the national C-ITS roadmap that Action 2.1 is now developing.[11] Action 2.1 requires the development and publication of a national C-ITS roadmap enabling a staged, nationally consistent approach. Lastly, Action 2.5 sets out a mechanism to identify impediments to implementing the eCall in-vehicle automated crash notification system in Australia.[12]
Are there any recent updates in the 5G space related to autonomous vehicles?
4.
[13] Ben Allen (n 7)
As noted above, the AVSL was originally expected to be in force by 2026, but that timeline was not achieved. The CAV Action Plan represents Australia's current legislative roadmap, designed to complete the end-to-end AV regulatory framework and enable safe commercial deployment by 2027. It sets out a comprehensive program of legislative and policy work at Commonwealth, state and territory level, and is the most significant structural step forward in Australia's AV regulatory journey to date. Once in effect, the AVSL will place responsibility for AV safety on a corporation, the Automated Driving System Entity (ADSE), rather than a human driver. The AVSL will be supported by complementary Commonwealth and state legislative changes, forming an end-to-end framework covering first supply, ADSE capability, on-road safety duties and user responsibilities ADSE status is not determined solely by who built the technology; the key question is whether an organization has sufficient influence over safety-critical decisions. Many businesses, including software providers, fleet operators and remote monitoring providers, will need to assess their position carefully in light of this framework.[13]
Have any new laws/regulations been passed within the last year that have affected the deployment of autonomous vehicles?
5.
Key CAV Action Plan actions timed for 2026 include: Action 2.3(b): the development of a business case for a national C-ITS security credential management system, which will provide a coordinated national approach to authenticating and securing communications between connected vehicles and roadside infrastructure, a foundational requirement before large-scale C-ITS deployment can proceed; Action 2.4: continued monitoring of radio frequency spectrum arrangements for C-ITS to ensure they remain fit for purpose and appropriately aligned with key international markets, covering both short-range and long-range connectivity technologies; Action 3.6: investigation of appropriate in-vehicle alert standards, including visual, auditory and haptic (touch-based) systems that communicate the operational status of automated driving systems to occupants and guide appropriate human responses, which is particularly important for ensuring safety at the boundary between human-driven and automated modes; Action 3.9: investigation of a national coordination role for Mobility as a Service (MaaS) platform, considering how automated vehicles can be integrated across transport modes and providers to improve accessibility, efficiency and service delivery, particularly for users who do not or cannot drive; and Action 3.10: identification of strategies for reducing greenhouse gas emissions through optimal CAV deployment, recognizing that the efficiency gains from connected and automated driving could contribute meaningfully to Australia's broader transport emissions reduction objectives. Work has also commenced on updated state and territory road traffic laws to regulate human users of AVs. Action 1.8 captures the National Transport Commission's regular readiness assessments, timed for 2025 and 2027.
What advancements in autonomous vehicles should we expect from your country in 2026?
6.
Australia's geography and industrial profile make it one of the most compelling environments in the world for commercial AV adoption beyond the passenger vehicle context. The dispersed population centres, significant physical separation between major metropolitan areas, and the scale and remoteness of Australia's mining operations, agricultural landholdings and freight corridors create both a strong economic case and a practical necessity for automation. Australia's mining industry, concentrated in remote regions of Western Australia and Queensland, has long operated at the frontier of heavy vehicle automation. Its agricultural sector faces comparable pressures: chronic labor shortages, vast broad acre properties spanning hundreds of thousands of hectares, and the imperative to maximize agricultural output with fewer resources have similarly driven substantial investment in autonomous and semi-autonomous machinery. The logistics and freight industries face a parallel challenge. Servicing widely separated urban markets across the country demands supply chain solutions that human-operated transport alone struggles to deliver at scale. Commercial AV deployment in Australia is therefore not simply a question of technological ambition, but increasingly a matter of industrial necessity. The following examples illustrate how that necessity is already translating into real-world deployment across four key sectors: Road and Freight Trucking A multinational vehicle manufacturer and an Australian motorway operator commenced a trial in August 2025 of C-ITS technology on Australian motorways. The motorway operator has also partnered with an international autonomous driving software company, building on a 2022 self-driving truck trial on major Melbourne motorways, to explore how Level 4 autonomous driving technology combined with smart road infrastructure could make heavy freight trucking safer and more efficient. Mining There are already automated trucks operating on mine sites in north-west Western Australia. A multinational truck manufacturer is preparing to launch what is described as the world's first fleet of autonomous in-pit mining trucks at an Australian manganese mine, where autonomous rigid tippers will transport ore across the site. Farming Australia's agricultural sector has emerged as a significant arena for autonomous vehicle deployment, driven by acute labor shortages and the operational demands of vast broad acre properties. GPS-guided autonomous tractors, self-driving seeders and robotic harvesting machinery operate increasingly without human intervention on large-scale grain and cotton farms, with major equipment manufacturers having introduced autonomous and semi-autonomous combines to the Australian market and the Government's On Farm Connectivity Program, providing dedicated funding support for the connectivity infrastructure on which these systems depend. Warehousing and Industrial Logistics Autonomous vehicle technology is increasingly being adopted in controlled, off-road environments such as warehouses, factory floors and mine sites. Australian operators are actively deploying purpose-built autonomous vehicles in industrial and logistics settings. Because these vehicles operate off public roadways, they fall outside the scope of the AVSL framework and are instead governed by existing workplace health and safety obligations.
Aside from robotaxis, do you see any other developments in the commercial space regarding AVs, including farming, commercial trucking, food delivery, etc.?
7.
[14] Department of Industry, Science and Resources (Cth), Safe and Responsible AI in Australia: Australian Government's Interim Response (Report, 2024). [15] Safe and Responsible AI in Australia: Australian Government's Interim Response (n 16); Ben Allen (n 7). [16] 2024–27 National Connected and Automated Vehicle (CAV) Action Plan (n 1).
Australia has not yet enacted standalone AI-specific legislation. However, the interaction between AI governance and the AV regulatory framework is a live and developing area, with the government having classified AI-enabled vehicles as a higher-risk application. In the current absence of dedicated AI legislation, AV operators must currently navigate a combination of existing sector-specific laws and the AVSL framework. Further clarity is expected as Action 3.1 of the CAV Action Plan progresses. It is also important to note that the artificial intelligence underpinning autonomous vehicles is fundamentally different in character from the large language model AI that has attracted significant public and regulatory attention in recent years. AV artificial intelligence does not operate primarily by processing and generating natural language. Rather, it is a complex integration of neural networks, machine learning algorithms and real-time sensor data-processing models that must interpret dynamic physical environments, make safety-critical decisions within milliseconds and interact with unpredictable road conditions at speed. This technical distinction has direct regulatory implications. Legislative frameworks designed with generative AI in mind may not be well-suited to addressing the specific risk profile of AV systems, and any future Australian AI regulation will need to account for this difference if it is to engage meaningfully with the autonomous vehicle sector. AVs may be enabled by AI, and the government has classified this as a higher-risk application. AI-related concerns relevant to AV deployment are addressed through the existing cross-cutting legal framework and the holistic law review contemplated by Action 3.1.[14] Safety and privacy compliance can no longer be treated as separate workstreams. The ADSE is the organization legally responsible for the safety of an automated driving system throughout its operational life, and ADSE status hinges on whether an organization has sufficient influence over safety-critical decisions, making governance of the underlying AI system a central question under the AVSL framework.[15] Australia is participating in United Nations work to develop international AV standards, which will form the basis of ADRs. This is the primary mechanism through which international AI governance principles may influence the Australian AV regulatory framework in the absence of domestic AI-specific legislation. As of early 2026, mandatory AI-specific regulation had not been enacted. Whether any future AI regulation will intersect with the AVSL framework remains to be resolved under Action 3.1 of the CAV Action Plan.[16]
Have any AI laws impacted deployment?
8.
Australia's approach to AV regulation is, in substance, a deliberate "wait and see" posture. Rather than moving to enact prescriptive AV-specific legislation ahead of widespread deployment, the government has chosen to observe how the technology matures and comparable jurisdictions, most notably the United States, resolve the legal, technical and policy tensions that AV integration presents. The primary evidence for this characterization is the trajectory of the AVSL framework itself. Despite years of consultation and the publication of successive action plans, the framework has not been fully operationalized, and the Government's own 2026 implementation milestone has not been met. Australia's regulatory position is not, however, shaped by legislation alone. The pace and direction of AV integration in Australia is also being guided by domestic industrial needs, as evidenced by the rapid uptake of autonomous vehicle technology in the mining, agricultural and logistics sectors, and by shifting consumer trends, including growing familiarity with ADAS features and increasing public exposure to AV technology through internationally available consumer vehicles. This cautious incrementalism may reflect a pragmatic recognition that legislating ahead of the technology carries its own risks. It also means, however, that Australia enters this period of rapid commercial AV development without the regulatory certainty that both industry and the public will ultimately require.
Conclusion
SAE Level
Name
Description
Example
0
No automation
Features limited to providing warnings and momentary assistance. The human driver performs all aspects of the driving task.
Lane departure warning
1
Driver assistance
Features provide steering or brake/acceleration support to the driver.
Adaptive cruise control
2
Partial driving automation
Features provide a combination of steering and brake/acceleration support. The human driver remains responsible for all other aspects of the driving task.
Lane keeping and adaptive cruise control simultaneously
3
Conditional driving automation
The vehicle can drive itself under limited conditions, but a user must be receptive to requests to intervene when required.
Traffic jam chauffeur
4
High-driving automation
The vehicle can drive itself under limited conditions with no expectation that a user will need to intervene.
Driverless taxi operating within a defined geographic area
5
Full driving
The vehicle can drive itself under all conditions without any expectation that a user will need to intervene.
Driverless taxi without operational limitations
Many vehicles on Australian roads already have Advanced Driver Assistance Systems (ADAS) such as lane keeping assist, assisted breaking and adaptive cruise control, but these do not make a vehicle "automated" in the regulatory sense. The critical legal threshold is the transition from Level 2 to Level 3, at which point the automated driving system takes over and human responsibility begins to diminish. Current Australian Road Rules were developed for SAE Levels 0-2, and the Automated Vehicle Program is focused on updating them to accommodate Levels 3 and above. Achieving this requires not only Commonwealth legislative action, but also parallel amendments to road traffic laws in each state and territory, making the harmonization process both technically complex and politically challenging. The CAV Action Plan addresses this through: Action 1.4, requiring states and territories to update their road traffic laws and enforcement activities to support automated vehicle operations; Action 1.5, monitoring motor accident injury insurance scheme changes across jurisdictions to align with the AVSL; and Action 3.1, holistically assessing existing Commonwealth, state and territory laws and identifying legislative gaps that could impede nationally consistent deployment. Vehicle connectivity standards will be translated to Australian Design Rules (ADRs) as they are finalized internationally, with a focus on ensuring imported vehicles can use Australia's C-ITS systems.
Back to top of page
While 2025 saw few legal developments related to autonomous vehicles (AVs), it was notable for increased implementation of automated commercial transport vehicles, among other things. Ontario launched a new pilot project to support the development of commercial AVs (see Question 5 below), with much of the push coming from private sector initiatives. For example, the Canadian Automobile Vehicle Initiative (CAVI) launched Phase 2 of its Trans-Canada Autonomous Truck Demonstration Project.[1] The project aims to have a driverless tractor-trailer complete a journey from Halifax to Vancouver in 2028. Phase 2 involves the development of detailed plans by various functional committees organized by CAVI. Notably, Canada’s largest grocery chain, Loblaw, entered into a contract with Gatik, a leader in autonomous freight for regional logistics networks, to enhance regional distribution in the Greater Toronto Area with self-driving trucks.[2] Loblaw claims that the multi-year growth agreement signifies the largest planned implementation of autonomous trucks in North America. Finally, Toronto-based Waabi Innovation Inc., an autonomous vehicle start-up, announced in early 2026 that it raised $750 million to continue its deployment of autonomous long-haul trucking and robotaxis – one of the largest venture capital financings in Canadian history.[3] Waabi’s operations are and will continue to be in partnership with Uber, both on a robotaxi and freight level.
What are the latest autonomous vehicle developments in 2025?
[1] CAVI, CAV Update (November 24, 2025). [2] Loblaw Companies Limited, Gatik and Loblaw Ink 5-year Expansion Deal to Scale AI-Powered Autonomous Trucking Solution Across Regional Distribution Networks in the Greater Toronto Area (September 23, 2025), Loblaw to deploy 50 fully autonomous Gatik trucks in Toronto [3] The Globe and Mail, Waabi launching robotaxi fleet with Uber after raising US$750-million (January 28, 2026).
None of the legal developments in 2025 related to AVs pose new roadblocks. While Canada’s federal government released guidance in the Safety Framework for Connected and Automated Vehicles 2.0 (the Safety Framework 2.0)[4] , this document merely presents existing law and non-regulatory guidance. Outside the regulatory sphere, organized labor may pose a roadblock to AV development in some circumstances. The Port of Prince Rupert in British Columbia on the Pacific Coast (the third-largest port in Canada) anticipates a driver shortage as early as 2026, while at the same time expecting a rise in cargo due to expansion projects set to be completed in 2026 and 2027.[5] The Port has started testing self-driving trucks to address the anticipated gap between drivers and cargo. In response, the union representing workers at the Port stated that it is “deadset opposed to the testing,” arguing that it could lead to job losses and safety issues. A continuous challenge for the development of AVs in Canada is regulatory differences across the 10 provinces and three territories. Among Canada’s most populous provinces, there remains significant divergence. While Ontario[6] and Quebec[7] allow SAE Level 3 vehicles on the road, British Columbia does not permit public operation of SAE Level 3 or higher vehicles.[8]
Canada’s cybersecurity and privacy landscape for autonomous and connected vehicles (together, CAVs) has advanced meaningfully over the past year, with developments at both the federal and provincial levels. Federally, Ottawa has reintroduced comprehensive critical infrastructure cybersecurity legislation as Bill C‑8 (An Act respecting cyber security), which would establish the Critical Cyber Systems Protection Act (CCSPA). If passed, CCSPA would impose mandatory cybersecurity program, supply‑chain and third-party risk management, and incident reporting requirements on operators of “vital” systems and services, including federally regulated road transportation systems. At the same time, broader private sector privacy reform is pending following the lapse of Bill C‑27 in January 2025. The federal government has signaled that a new private sector privacy bill is expected in 2026, but until then, PIPEDA continues to govern privacy considerations related to in‑vehicle data, telematics and CAV pilots. At the same time, Transport Canada has continued to strengthen the national cybersecurity baseline for vehicles. Its Vehicle Cyber Security Strategy and Vehicle Cyber Security Guidance provide a foundation for strengthening CAV cyber security in Canada by outlining best practices on governance, layered defenses, supply‑chain security and incident response. Canada’s Safety Framework for Connected and Automated Vehicles 2.0 reinforces these principles and directs stakeholders to assessment tools such as the Vehicle Cyber Security Assessment Tool and ITS self‑assessments. Updated federal testing guidance, the Guidelines for Testing Automated Driving Systems (v2.0), also emphasizes cybersecurity and incident‑preparedness obligations for CAV trials, including coordination with federal and provincial privacy regulators. Research capacity expanded as well: in 2025, researchers from the National Research Council partnered with researchers from the University of Waterloo and policymakers from Transport Canada to virtually stress‑test AI‑driven CAV decision‑making against cyberattacks, helping translate policy expectations into engineering practice without putting real vehicles at risk. At the provincial level, Ontario passed Bill 194 (Strengthening Cyber Security and Building Trust in the Public Sector Act, 2024), Schedule 1, which enacted the Enhancing Digital Security and Trust Act (EDSTA). The EDSTA creates new cybersecurity and breach reporting expectations, and addresses obligations related to the use of artificial intelligence systems by public‑sector entities governed under the Freedom of Information and Protection of Privacy Act (FIPPA) and its municipal counterpart (MFIPPA). Bill 194 also amended FIPPA to, among other things, require that institutions conduct privacy impact assessments respecting any personal information that they intend to collect. Alberta also saw significant legislative reform with the replacement of the province’s Freedom of Information and Privacy Act with a new public sector privacy law, the Protection of Privacy Act (POPA), which came into effect in June 2025. Similar to the obligations imposed under Ontario’s Bill 194, POPA has enhanced breach reporting and privacy impact assessment obligations. These new obligations are expected to impact the use of CAVs by public-sector institutions in both provinces. Canadian privacy regulators have also increased scrutiny of in‑vehicle surveillance and roadside data collection, which underscores the need for privacy impact assessments, clear notices and data‑minimization practices for camera‑, sensor‑ and biometric‑enabled CAV deployments. Together, these developments demonstrate a continued shift toward structured cybersecurity and privacy obligations, setting a foundation for the secure, privacy‑protective deployment of CAVs across Canada.
Advanced 5G features, such as ultra-reliable low-latency connections, massive machine-to-machine (M2M) communication and network slicing are key enablers of the development of autonomous vehicles. While initial deployments of 5G networks in Canada have focused on coverage,[9] current and future deployments will increasingly focus on generating higher data transfer rates, improved connectivity and higher system capacity. Canada’s spectrum regulator, Innovation, Science and Economic Development (ISED), anticipates that demand for commercial mobile services will continue to grow, driven in part by demand for and take-up of M2M connectivity by business users for innovative applications in areas such as health services, autonomous vehicles and connected cities. At present, ISED believes that the spectrum released and planned for release in Canada is adequate to meet anticipated demand. The CRTC and the telecommunications industry in Canada have also done the necessary work to implement telephone numbering resource policies and procedures to prepare for the proliferation of devices in the Internet of Things (IoT) universe. In support of the autonomous vehicle agenda in Canada, ISED has specifically encouraged the creation and development of the Intelligent Transportation System (ITS) industry vertical through its spectrum management and telecommunications powers. ITS will connect people, vehicles, roadside infrastructure and wide area networks to ensure safer and more efficient road transport and transportation infrastructure. For example, one feature of safety-related ITS applications enables drivers to quickly receive information about imminent obstructions or dangers, thereby allowing them to take action to prevent potential accidents. Since 2023, with respect to ITS in Canada, ISED has: In addition to the foregoing, ISED has established a non-competitive local licensing (NCLL) framework and process to promote the growth of innovative business models and enable small operators and businesses to gain easier access to wireless services and enhance connectivity in Canada. The NCLL may facilitate improved productivity and worker safety by enabling the establishment of wireless networks to enable remote equipment operation and monitoring, including autonomous vehicles and safety systems.
[10] Canada, Canada’s Safety Framework for Connected and Automated Vehicles 2.0 (February 11, 2025), section 1.4. [11] Ontario, O.Reg 268/25: PILOT PROJECT - AUTOMATED VEHICLES (November 27, 2025). [12] Ontario, O. Reg. 161/25: PILOT PROJECT - AUTOMATED COMMERCIAL MOTOR VEHICLES (August 1, 2025). [13] Quebec, Autonomous vehicles on Québec roads (July 18, 2025).
The federal government released the Safety Framework 2.0 in February 2025.[10] As mentioned in response to Question 2, this document mostly describes existing federal AV laws and regulations. Still, it provides some useful guidance with respect to aspects of the federal government’s jurisdiction over AV issues. The federal government has jurisdiction over automobile safety standards and specifications, among other things, and the Safety Framework 2.0 describes how flexibility in existing regulations allows for the development of AVs that might otherwise not meet federal safety standards. First, a person or company may temporarily import a non-compliant vehicle for testing purposes via a declaration system facilitated by Transport Canada – the federal ministry primarily responsible for AV matters. Second, Canada’s Minister of Transportation may grant exemptions from safety standards for a vehicle for a specified period to allow for technological development, so long as the exemption would not substantially diminish the overall safety performance of the vehicle. At the provincial level, Ontario extended the lifespan of an existing AV pilot project and introduced a new one. In November 2025, the Ontario government enacted Regulation 268/25, which extended the life of Ontario’s general AV pilot project to October 13, 2027.[11] The project would have otherwise ended January 1, 2026. Separately, the Ontario government created Regulation 161/25 on August 1, 2025, a pilot project for commercial AVs.[12] The new project invites applications for the testing of commercial motor vehicles at SAE Levels 3-5, provides some requirements of the vehicles tested, and imposes duties on drivers, among other things. The Quebec government released an AV guidance note on July 18, 2025, describing the current law regarding such vehicles in Quebec.[13] The note specifies that Quebec permits SAE Level 3 vehicles, and that Level 4-5 vehicles are allowed if they are part of a Quebec pilot project.
Both the public and private sectors will spur AV advancements in 2026. From a public perspective, Canada’s Safety Framework 2.0 provides some indication of public opinion regarding AVs in Canada, and how the federal government intends to continue its support of development and implementation.[14] Safety Framework 2.0 explains that Transport Canada conducted surveys in 2019 and 2021, which found that Canadians are mostly unfamiliar with AVs and are skeptical of such systems. Additionally, the surveys note confusion regarding the terms used for AV systems, given that manufacturers may use different names for the same systems. Safety Framework 2.0 notes that the federal government is working with stakeholders to promote the standardization of autonomous driving system terminology. This suggests that 2026 may see a greater push for public education regarding AV technologies. Moreover, Safety Framework 2.0 indicates research priorities for the federal government, which in turn will shape development of AVs in Canada in 2026 and beyond.[15] Safety Framework 2.0 identifies the following research priorities: Automated driving system safety testing, human factors research, to better understand the safety implications of human interactions with AV systems, crash avoidance, vehicle communication, and cooperative truck platooning systems, which involves developing systems that enable multiple autonomous trucks to closely follow each other. At the provincial level, Ontario will likely continue to invest AV technology projects through the Ontario Vehicle Innovation Network (OVIN). For example, 2026 may see the fruits of OVIN’s partnership with the Toronto Transit Commission (TTC) – Toronto’s public transit operator. OVIN announced a partnership with TTC involving grants to companies of up to $100,000 to pilot new technologies to address Toronto’s urban mobility challenges.[16] Such technologies will likely include AVs. From the private sector, Canada may see a stronger push from businesses to implement robotaxis in 2026 and beyond. Whether these can be accommodated within the existing regulatory framework (including pilot programs in Ontario and elsewhere) raises an interesting question. Waymo, a subsidiary of Alphabet, which develops fully self-driving vehicles for ride-hailing, has registered as a lobbyist in various Canadian jurisdictions.[17] This was a late development in 2025, and it is yet to be seen how such lobbying efforts will play out. While the lobbyist registry for the federal government indicates no activity as of writing, the registry in British Columbia indicates that Waymo has made 36 activity reports since registration on July 30, 2025.[18] Waymo already operates in cities including San Francisco, Phoenix and Los Angeles. In addition to the actual testing and deployment of autonomous vehicles, 2026 may see further development of the technological infrastructure to support AV use. The CEO of Toronto-based Pudocity explained in a 2025 interview that the company focuses on “orchestration-as-a-service” to optimize the coordination of AV fleets in a region. When AV ride-hailing services enter Canada, we may see a relationship between regional transport authorities and companies such as Pudocity to ensure a smooth simultaneous operation of several independent AV ride-hailing companies.
[33] Parliament of Canada, Bill C-27. [34] Canada, 2025 Budget, at page 92. [35] Canada, 2025 Budget, at page 92.
[4] Canada, Canada’s Safety Framework for Connected and Automated Vehicles 2.0 (February 11, 2025). [5] CBC, Unions oppose B.C. port's plans to test self-driving trucks (May 30, 2025). [6] Ontario, Connected and automated vehicles (January 25, 2022). [7] Quebec, In an Autonomous Vehicle (July 18, 2025). [8] British Columbia, Automated (self-driving) vehicles (April 4, 2024).
Designated the 5895-5925 MHz band for ITS devices and specifically mandated the use of Cellular Vehicle-to-Everything (C-V2X) technology in Canada, which operates using on-board and roadside units that can directly communicate with other connected vehicles, road users and roadside infrastructure. ISED’s decisions related to C-V2X are designed to harmonize the Canadian ITS ecosystem with that adopted by the FCC in the United States in order to promote cross-border interoperability of connected vehicles. Allowed on-board ITS units to operate on a license-exempt (no protection, no interference) basis and developed standards to enable license-exempt operation of certified ITS on-board units. Determined that it will develop a licensing framework for roadside ITS units through a future consultation process.
[14] Canada, Canada’s Safety Framework for Connected and Automated Vehicles 2.0 (February 11, 2025), section 2.2. [15] Canada, Canada’s Safety Framework for Connected and Automated Vehicles 2.0 (February 11, 2025), section 4.2. [16] Ontario, OVIN and the TTC to Pilot New Made-in-Ontario Transit Technologies (November 27, 2025). [17] CBC, Could Torontonians soon ride self-driving taxis? That’s Waymo’s plan (December 22, 2025); Business in Vancouver, Robotaxi giant Waymo lobbying B.C. for changes to ban on driverless vehicles (December 3, 2025). [18] Canada, Registry of Lobbyists, Waymo LLC / Garry Keller, Consultant (October 27, 2025); British Columbia, Officer of the Registrar of Lobbyists, Waymo LLC / Trevor Kramer, Consultant Lobbyist (July 30, 2025).
[19] CAVI, CAV Update (November 24, 2025). [20] CBC, Unions oppose B.C. port's plans to test self-driving trucks (May 30, 2025). [21] Montreal Economic Institute, Canadian Ports Have a Productivity Problem (February 2025). [22] Global Sensor Systems, AI-Based Fleet Safety Innovation by Global Sensor Systems and OMS Express receives support (August 13, 2025); Joyride, Joyride officially unveils Ontario’s first IoT-connected GEM vehicle (August 13, 2025); Dejero, Ontario Invests in Dejero to Advance Next-Generation Fleet Connectivity Innovation (June 26, 2025); [23] CHCH, Canada in robot trucking vanguard, with one driverless semi already here (April 21, 2025). [24] CBC, Food delivery robots in Markham test appetite for high-tech takeout (June 3, 2025). [25] CBC, A pilot project is bringing self-driving robots to Toronto — sparking concerns from one city councillor (May 4, 2025). [26] Global News, Magna’s automated vehicles faced construction and red light issues: city report (December 8, 2025). [27] Aurrigo, Aurrigo launches Canada’s first all-season, medium-speed autonomous shuttle in Kanata (August 18, 2025). [28] CBC, Kitchener partners with local startup to lay groundwork for autonomous shuttles (July 15, 2025). [29] GeoMate, OVIN supports partnership between Kitchener and GeoMate for dynamic HD maps for autonomous driving applications (January 2025). [30] Kratos, Kratos Unmanned Systems Demonstrates Leader Follower Platooning Technology in Quebec's Forestry Sector in a Timber Hauling Operation (January 27, 2025). [31] Suncor, The Future of Canadian Mining is Autonomous — and Suncor is Leading the Way (June 30, 2025). [32] Electric Autonomy Canada, Autonomous battery-electric loaders now in operation at Canadian gold mine (January 8, 2025).
[9] According to the latest data published by the CRTC, 99.5% of Canadians had access to LTE coverage by the end of 2024. Coverage was lower for rural populations (97.6%): CRTC, Communications Market Report, Reported Mobile Coverage Across Canada, online: https://crtc.gc.ca/eng/publications/reports/PolicyMonitoring/mob.htm
i.
ii.
iii.
Federal/national As mentioned in response to Question 1 above, CAVI is implementing Phase 2 of its Trans-Canada Autonomous Truck Demonstration Project to demonstrate a fully autonomous, cross-Canada tractor-trailer.[19] If successful, the demonstration may spur further adoption of commercial AVs for the purpose of transporting goods across provincial borders. As mentioned in response to Question 2 above, the Port of Prince Rupert in British Columbia is testing deployment of self-driving trucks to address an anticipated shortage of workers and increased cargo demand.[20] Indeed, 2026 and beyond may see such testing at ports across Canada. A brief 2025 paper from the Montreal Economic Institute, a think tank, explains that Canadian ports have poor productivity relative to their international peers, and argues that automation is a promising avenue for increasing efficiency.[21] Ontario In addition to Loblaw’s expansion of self-driving trucks and Waabi’s significant financing announcement mentioned above in Question 1, Ontario saw several small-scale developments in the commercial and public transportation implementation of AVs in 2025. The Ontario government, through OVIN, provided financial support to various commercial AV technology developments led by small and medium-sized businesses.[22] Canadian Tire, a large Canadian consumer goods retailer, partnered with NuPort Robotics to pilot the use of an autonomous 18-wheeler to shuttle goods within a distribution center north of Toronto.[23] In Markham, Ontario, Real Life Robotics, an automation company, partnered with Skip, a food delivery service, to pilot autonomous food delivery robots that operate exclusively on the sidewalk.[24] Magna International, a Canadian auto parts manufacturer, launched a pilot project of self-driving delivery robots in Toronto.[25] The pilot involves 20 self-driving vehicles under the observation of humans on streets where the speed limits are 40km/h or lower. Results of the pilot project have been mixed: the City of Toronto says it has received several unsolicited reports of the delivery robots experiencing confusion at construction sites and difficulty turning right at a red light.[26] Aurrigo, a developer of AV technology, is deploying nine-seater passenger vehicles to aid transportation across eight stops in the Kanata North Technology Park just outside of Ottawa.[27] The project is notable in that the shuttles will have to operate in the challenging winter conditions experienced in the Ottawa region. The City of Kitchener partnered with GeoMate, a local startup, to create an artificial intelligence-driven map to aid the potential use of AV shuttles in Kitchener’s downtown.[28] The partnership involves funding from the Ontario government’s Vehicle Innovation Network, an $8 million investment announced in 2024. OVIN has also provided support for GeoMate’s mapping software.[29] Quebec Quebec saw a notable commercial AV development in its prominent forestry industry led by the private sector. Kratos Defense and Security Solutions, a technology company in defense and national security markets, successfully executed a multi-week demonstration of a truck platooning system in northern Quebec.[30] In partnership with FPInnovations, a private not-for-profit research and development organization focused on Canada’s forestry industry, Kratos deployed two tractor trailers carrying lumber along supply chain routes in Quebec, with one human-driven leader truck to collect data and transmit the data to a self-driving follower truck. Alberta Alberta’s energy industry also saw continued implementation of AVs. Suncor, a global energy company based in Calgary, issued a press release on June 30, 2025, describing its ongoing implementation of ultra-class autonomous haul trucks.[31] Suncor began rollout of the fleet in 2018, growing to 15 haul trucks in 2024. It noted that ongoing improvements to truck performance on different road and seasonal configurations have allowed for continued deployment into 2025, with the fleet now numbering 120 across two integrated mines near Fort McMurray. British Columbia The mining industry in British Columbia saw one notable development regarding AVs in 2025. Sandvik, a Canadian mining company, deployed two 18,000-kilogram autonomous electric loaders to its New Afton gold mine outside of Kamloops, British Columbia.[32] Sandvik anticipates that the deployment will enable increased efficiency and reduced heat, noise and carbon emissions.
The 2025 iteration of this report discussed how the federal government’s Artificial Intelligence and Data Act, introduced in 2022 as part of Bill C-27, might impact the deployment of AVs once enacted. The law is designed to specifically regulate “high-impact” artificial intelligence (AI) systems, including systems critical to health and safety such as AV systems. As of this writing, Parliament will need to reintroduce Bill C-27 if it is to ever become law.[33] Parliament’s last consideration of Bill C-27 was in a meeting of the Standing Committee on Industry and Technology on September 26, 2024. As Bill C-27 was not passed during that session of Parliament, the law “died,”, such that its future depends on whether the current or a subsequent Parliament reintroduces the law. That said, several elements of Canada’s November 2025 Budget (2025 Budget) target the development of AI technologies, which in turn may support the deployment of AVs in Canada. Two aspects of the 2025 budget are notable in this respect. First, the 2025 budget commits $925.6 million over five years to build large-scale sovereign compute capacity. It explains that such investments will help allow for increased automation throughout the economy.[34] Second, the 2025 budget announced that the Minister of Artificial Intelligence and Digital Innovation will engage with industry to identify promising AI infrastructure projects and enter into Memoranda of Understanding with those projects.[35] While “infrastructure” does not necessarily contemplate AVs, the deployment of AVs will likely involve AI-powered infrastructure akin to the robotaxi orchestration services discussed in response to Question 6 above.
< ARIA label should be applied here
E.g. A button
Accessibility Checklist Please make sure that Accessibility is turned on in the experience settings Sections logically grouped with HTML tags Sections should be logically grouped and named. These should have an HTML section tag added based on the content/function. E.g. for mastheads, / for content groups.
2025 is a pivotal year for China’s autonomous driving technology, transitioning from demonstration applications to large-scale commercialization and compliance. The following are the latest core developments in China’s driverless sector in 2025: First, China's autonomous driving industry witnessed a landmark policy breakthrough in 2025. In December 2025, the Ministry of Industry and Information Technology officially issued the first batch of approval permits for L3 conditional autonomous driving vehicles. The first approved models include a pure electric sedan from Changan Automobile and related models from BAIC Arcfox. This confirms that passenger vehicles equipped with L3 autonomous driving capabilities have obtained national-level official approval in specific scenarios such as highways and urban expressways. They have officially shifted from experimental testing to legal mass production and commercial deployment.[1] Secondly, robotaxi commercial operations have been launched with intense industry competition. In the mobility service sector, Apollo Go maintained its leading position in 2025. Its autonomous driving mileage has exceeded 240 million kilometers. It has provided more than 17 million mobility services globally. Apollo Go has achieved regular operations in multiple cities, including Wuhan, Shanghai and Guangzhou.[2] Meanwhile, Pony.ai and WeRide both completed filings with the China Securities Regulatory Commission for IPO preparations. They have also actively expanded into overseas markets such as the Middle East and Europe.[3] By 2025, robotaxis were no longer limited to closed test areas. Beijing, Guangzhou and other cities have widely launched commercial paid pilot programs with no one in the driver’s seat.[4] Meanwhile, cities are gradually popularizing NOA and the End-to-End architecture. Technically, 2025 is the year when urban NOA (Navigation On Autopilot) has been fully rolled out. According to a report by the China Association of Automobile Manufacturers (CAAM), the adoption volume of urban NOA from third-party suppliers, including Momenta, has increased significantly, capturing an important share of the market.[5] Leading domestic automakers such as Huawei and Xpeng both accelerated the shift to the End-to-End neural network architecture in 2025. They used large models to raise the upper limit of vehicles’ ability to handle complex traffic scenarios.[6] In addition, the commercialization of low-speed autonomous driving has reached a key stage in logistics and sanitation scenarios. Policies in various provinces and cities are promoting the gradual deployment of autonomous delivery vehicles.[7] As technologies are deployed, laws and regulations have provided clearer answers on liability attribution. New standards issued in 2025 make it clear that, if a traffic accident occurs while the autonomous driving system is activated, liability shall in principle be borne by the automaker or operator.[8] In summary, driven by policies, technologies and capital, China's autonomous driving industry achieved a critical leap from "demonstration applications" to "large-scale compliant implementation" in 2025. The approval of L3 vehicles was launched, robotaxi commercial operations became widespread and End-to-End technology became mainstream. The industrial ecosystem has become increasingly mature.
[1] "Breaking the ice for L3-level access: Commercial implementation of autonomous driving is entering the fast lane," Beijing Business Daily [2] "Baidu Apollo Go operates 250,000 times per week, with L4 autonomous driving on par with Waymo" [3] 2025-10-14 19:21 Official account of Cailian Press [4] "Can you sit with peace of mind with Robotaxi now?" [5] "Dual Strengths in Intelligent Driving" Landscape! Huawei Momenta together accounts for 80% of the third-party city NOA market" [6] "Huawei, NIO, XPeng, and Li Auto have entered the market one after another, with mass vehicle adoption expected by 2025. The inaugural year of "end-to-end" autonomous driving large models is approaching" [7] "The era of low-speed autonomous driving is accelerating," China Securities Journal [8] "Notice from the Ministry of Industry and Information Technology and the State Administration for Market Regulation on Further Strengthening the Management of Intelligent Connected Vehicle Product Entry, Recall, and Online Software Upgrades"
Despite China’s remarkable progress in 2025, it is facing a series of severe challenges on the path to full commercialization. Below are the core obstacles to the development of China’s autonomous driving in 2025. First, on the cost side, hardware premiums exist while commercial profitability remains difficult. Although LiDAR prices have dropped substantially, the costs of vehicle redundancy systems and high-computing-power chips for true L4 autonomous driving are still high. Operating costs for robotaxis have decreased, but most platforms remain unprofitable due to vehicle depreciation and high expenditures on remote monitoring manpower.[9] In 2025, some intelligent driving suppliers encountered layoffs or restructuring due to capital chain ruptures. Intensified internal competition in the industry further raised the difficulty of achieving profitability.[10] Although many cities have issued L3 approval permits, policy fragmentation remains an obstacle to national promotion. Recognition standards and mutual recognition mechanisms for testing licenses for autonomous driving have not yet been fully unified across cities. For example, an autonomous delivery vehicle approved in Beijing may not be allowed to operate directly in the neighboring city of Langfang. Meanwhile, as geographic information data security has been elevated to a national strategic priority, high-definition maps and environmental data collected by autonomous vehicles are subject to strict regulation. In 2025, Chinese intelligent driving companies encountered unprecedented resistance in global expansion. Affected by geopolitics and data sovereignty laws, Chinese automakers face stringent data desensitization requirements and software-hardware supply chain reviews when entering North American or European markets. This has slowed the globalization process of China’s intelligent driving technologies.[11] While End-to-End large models have enhanced the smooth operation of the system, there are still shortcomings when the system handles China's complex traffic environments. For instance, sensor malfunctions can occur in extreme rain and snow weather.[12] In 2025, multiple misjudgment incidents of intelligent driving systems in long-tail scenarios triggered a public trust crisis over their safety, forcing regulatory authorities to step in and rectify irregularities in the industry.[13] Although laws have initially clarified that automakers are liable for L3-level autonomous driving accidents, practical technical disputes remain in insurance loss adjustment when defining whether the take-over delay is 0.5 seconds or 1 second based on black box data. Insurance companies have not yet developed a mature pricing model for autonomous driving-specific insurance policies. Excessively high premiums and potential legal disputes have left ordinary consumers with a wait-and-see attitude toward purchasing high-level intelligent driving models.[14]
On January 1, 2025, China's first mandatory national standard for automotive cybersecurity, Technical Requirements for Automotive Vehicle Information Security (GB 44495-2024), officially took effect. This standard requires all newly applied-for vehicle models to have the ability to detect and defend against cyber attacks, and to encrypt the transmission of key data. According to the records of the China National Standards Full-text Public System, this means that starting from 2025, autonomous driving vehicles that fail to meet the safety protection standards will not be granted market approval, marking that security defense has become a mandatory standard configuration.[15] Meanwhile, authorities have strengthened cybersecurity reviews for vehicles before leaving the factory and incorporated OTA (Over-the-Air) upgrades into the recall supervision system. In February 2025, the Ministry of Industry and Information Technology and the State Administration for Market Regulation jointly issued a notice requiring automakers to declare the detailed technical parameters of combined driving assistance systems. It also emphasized that OTA upgrades must not be used to conceal vehicle defects or evade recall liabilities. This means that any repair involving cybersecurity vulnerabilities is subject to traceability and evaluation by regulatory authorities.[16] In addition, cross-border data flow became a key regulatory focus in 2025. In June 2025, the Ministry of Industry and Information Technology and seven other ministries and commissions issued the Guidelines for the Security of Automotive Data Export (2025 Edition) (Draft for Comments), which clearly defined three types of automotive data export behaviors for the first time and set quantitative declaration thresholds: the provision of important data overseas, or the cumulative provision of personal information of more than one million people or sensitive personal information of more than 10,000 people, must be declared for security assessment. Among them, algorithms, training data and feature data in driving automation scenarios are clearly classified into the category of important data.[17] On July 23, 2025, the Ministry of Science and Technology issued the Ethical Guidelines for the R&D of Driving Automation Technology, which integrates data security and privacy protection throughout the entire technology life cycle. The guidelines explicitly require algorithm models to be clearly recorded and traceable, and a full life cycle privacy protection mechanism to be established, strictly abiding by the principles of "informed consent" and "minimum necessity". At the ethical level, it prohibits algorithms from being biased due to age, gender and other factors, and requires enterprises not to disseminate false information inconsistent with the actual functions of the system, and to clearly define product classification and usage conditions.[18] In summary, in 2025, by implementing the mandatory national standard for automotive information security, refining the quantitative thresholds for data export, strengthening the supervision of OTA recalls and issuing ethical guidelines, China has built a full-chain compliance system covering vehicle factory delivery, data flow, software upgrades and algorithm ethics.
Network and infrastructure construction[19] By 2025, with the evolution of 5G toward 5G-Advanced (5G-A, commonly referred to as ‘5.5G’), China's 5G network has entered a phase of deepened deployment. Multiple key cities have achieved 5G-A network coverage and commenced commercial rollout. Operators are accelerating their 5G-A rollouts, with plans to achieve comprehensive commercial deployment across the main urban areas of 39 key cities this year, alongside deployment in core zones of more than 300 cities. Compared to traditional 5G, 5G-A offers higher speeds, lower latency and enhanced integrated communication-sensing capabilities. These characteristics support the ultra-low latency and high-reliability communication demands of autonomous driving, such as real-time V2X traffic sharing and coordinated vehicle decision-making. Collaborative applications of vehicle-to-everything and intelligent transportation systems[20] During practical testing and deployment across multiple cities, 5G networks and vehicle-to-infrastructure (V2X) communications have begun to integrate significantly. Relevant reports indicate that 5G-A-based integrated communication and perception within urban road networks can achieve real-time vehicle tracking and high-precision coordination. This provides more reliable network support for autonomous vehicles' perception and decision-making in complex traffic environments. Experts contend that the high-speed, large-bandwidth and low-latency characteristics of 5G-A represent one of the key technological underpinnings for realizing future commercialized autonomous driving scenarios, particularly for motorway autonomous driving and autonomous driving in complex urban road conditions. Port implementation[21] In March 2025, Chongqing Guoyuan Port formally deployed six new-energy autonomous intelligent container trucks, establishing the Yangtze River's first mixed-operation scenario combining intelligent and manned container trucks in its upper reaches. Equipped with 5G antennas and LiDAR systems, these trucks utilize 5G+Beidou technology for precise positioning, handling approximately 200 TEUs daily. Compared to traditional operations, the two automated lines saved more than two million yuan annually, providing a replicable model for inland port automation transformation. L3 approval implementation[22] In December 2025, the Ministry of Industry and Information Technology granted the first batch of L3 conditional autonomous driving vehicle approvals. Changan and Arcfox each secured approval for one pure-electric model, tailored for urban congestion and motorway scenarios respectively, with pilot operations commencing in designated areas of Beijing and Chongqing. The transition from system-assisted to system-dominant control at Level 3 introduces new liability definitions, marking a pivotal milestone in China's progression from testing to commercial deployment of Level 3 autonomous driving. Concurrently, the industry's regulatory and standards framework is being refined.
[23] "Providing solid legal guarantees for promoting the application of autonomous driving" [24] https://www.spc.org.cn/online/0becacfd363c754f6869f970bc78ee5a.html [25] "L3-level autonomous driving models receive conditional access permits, marking a critical milestone in commercial implementation" [26] "On the first day of 2025, Beijing's new regulations mark the first year of scaling up autonomous driving" [27] "L3 autonomous driving launches simultaneously in Beijing and Wuhan! Automakers such as Li Auto and BYD have fully prepared for the competition" [28] "Over 50 Chinese Cities Have Brought Out Rules for Self-Driving, Smart Connected Vehicles"
In terms of national regulations, the revision of the Road Traffic Safety Law continues to progress. The Ministry of Public Security, in collaboration with multiple departments, is refining supporting regulations concerning the tiered management of autonomous driving and the determination of accident liability. This clarifies the legal boundaries between driver assistance and autonomous driving, providing overarching guidance for implementation across all regions.[23] In terms of national standards, the State Administration for Market Regulation and the Standardization Administration of China have issued GB/T 45312-2025 “Design and Operational Conditions for Autonomous Driving Systems in Intelligent Connected Vehicles,” establishing technical specifications for the development and testing of autonomous driving systems. Concurrently, the fourth batch of recommended national standards for 2025 has been announced, encompassing the formulation and revision of multiple autonomous driving-related standards, including in-vehicle millimeter-wave radar and perception function assessment testing equipment.[24] In terms of industry access, the Ministry of Industry and Information Technology issued China's first batch of conditional autonomous driving vehicle type approvals for Level 3 in 2025, explicitly permitting designated models to conduct commercial trials within specified zones in Beijing and Chongqing. Concurrently, a dynamic monitoring mechanism for intelligent connected vehicle access was established.[25] Locally, the Standing Committee of the Beijing Municipal People's Congress adopted the Beijing Autonomous Vehicles Regulations on December 31, 2024, with the regulations taking effect on April 1, 2025. The regulations comprehensively govern the application of autonomous vehicles on Beijing's roads, industrial development and safety oversight, explicitly supporting technological innovation, industrial advancement and demonstration activities within the autonomous vehicle sector.[26] The Wuhan Municipal Regulations on the Development and Promotion of Intelligent Connected Vehicles came into force on March 1, 2025. These regulations prioritize support for pilot demonstrations and commercial rollouts of intelligent connected vehicles within public transport, smart parking and intelligent logistics sectors, explicitly encouraging innovative business models and multi-scenario integrated applications.[27] Beyond the aforementioned cities, more than 50 localities nationwide – including Guangzhou and Shenzhen – have introduced regional regulations or policies specifically for autonomous or intelligent connected vehicles. These encompass multiple aspects such as road testing, demonstration applications and safety reporting disclosure, thereby providing foundational support for local autonomous driving development and policy innovation.[28]
In terms of technological innovation: By 2025, China's autonomous driving technology will continue advancing toward higher levels and broader applications. The overall market is projected to see more than two-thirds of new vehicles equipped with at least Level 2 autonomous driving capabilities, meaning vehicles can coordinate steering, acceleration and braking under specific conditions. This marks a significant milestone in the rapid adoption of consumer-grade intelligent driving technology.[29] Concurrently, China's automotive industry has made significant strides in the research, development and application of Level 3 and above autonomous driving technologies. Multiple domestic manufacturers and technology firms have advanced the integration and testing of advanced intelligent driving systems. It is anticipated that from 2025 onwards, a cohort of passenger vehicles equipped with Level 3 intelligent driving capabilities will enter mass production. This signifies that the genuinely meaningful user experience of “conditional autonomous driving” will be within reach.[30] In terms of industrial ecology: China's autonomous driving industry ecosystem has evolved from single-manufacturing to multi-faceted integration encompassing “perception chips + AI algorithms + vehicle-road coordination services.” For instance, the substantial reduction in costs for high-precision sensor technologies such as LiDAR and millimeter-wave radar has driven steady enhancements in autonomous driving perception capabilities.[31] Moreover, numerous intelligent driving solution providers are spearheading business model innovation through deep collaboration and ecosystem synergy. Consequently, China's autonomous driving systems are evolving from isolated vehicle-based perception toward intelligent mobility platforms that integrate vehicle-to-everything connectivity, edge computing and remote cloud services.[32] In terms of infrastructure construction: By 2025, China's framework for intelligent driving demonstration applications had gradually taken shape. Multiple national-level test zones and open-road test routes for intelligent connected vehicles have been established nationwide, with cumulative test roads exceeding tens of thousands of kilometers. This had accumulated extensive empirical experience for the validation and operation of autonomous driving technologies.[33] Regarding vehicle-road coordination infrastructure, several Chinese cities have accelerated the deployment of vehicle-to-everything (V2X) infrastructure, achieving the interconnection of intelligent transport units such as traffic lights and 5G networks.[34] In terms of regulations and standards: By 2025, China's standards framework for intelligent connected vehicles – encompassing safety, data sharing and technology access – will be progressively refined. At the national level, foundational specifications covering functional testing, safety assessments and data compliance requirements for autonomous vehicles have been established, providing legal and regulatory underpinnings for market deployment. Notably, Chinese regulatory authorities have progressively clarified supporting rules concerning autonomous vehicle market access, functional scope definitions and accident liability allocation. This development will help mitigate safety risks and bolster public confidence. [35] In terms of commercialization and market breakthrough: By mid-2025, several Chinese autonomous driving and intelligent mobility companies had scaled up deployment of robot cab services while actively exploring overseas expansion. For instance, Apollo Go, a leading domestic autonomous driving service platform, has cumulatively completed millions of passenger trips, demonstrating trends toward enhanced technological maturity and improved commercial operational capabilities.[36] Moreover, the launch of mass-produced models such as the Zeekr 9X, equipped with Level 3 autonomous driving capabilities, in the latter half of 2025 signifies that China's advanced autonomous driving products are entering the mainstream consumer market. [37] In terms of operational models’ innovation: Several autonomous driving service operators in China are exploring an “integrated service system for cross-scenario autonomous operation and smart mobility,” aiming to integrate autonomous vehicle services with traditional public transport, shared mobility and urban logistics. For instance, robotaxi services are expanding beyond on-demand city center hailing to encompass diverse operational scenarios such as airport transfers, corporate campus commuting and short-to-long-distance shuttle services. Furthermore, certain autonomous driving enterprises have proposed a novel operational model integrating AI dispatch systems, remote safety monitoring systems and intelligent transport platforms. They aim to achieve cross-city, multi-scenario operational capabilities by 2026. [38]
Line-haul freight (autonomous trucking) On the one hand, commercial operating mileage on the enterprise side continues to accumulate. Inceptio officially disclosed that the cumulative commercial operating mileage of its autonomous heavy-duty trucks has exceeded 200 million kilometers (as of its year-end 2024 disclosure point).[39] Pony.ai’s official materials also clearly state that its autonomous trucks have been delivered in batches, and that it has been approved to conduct L4 platooning tests in Beijing and Guangzhou (a “1+N” formation to reduce labor costs).[40] At the same time, a Guosen Securities research report shows that, as of April 10, 2025, Pony.ai was operating more than 190 intelligent trucks, with cumulative mileage of about 57 million kilometers and cumulative freight volume exceeding 926 million ton-kilometers.[41] On the other hand, autonomous trucking technology is also improving continuously. On July 2, 2025, at the 17th International Transportation Technology and Equipment Exhibition, JD Logistics officially launched its self-developed unmanned light truck product, the JD Logistics VAN, marking another important innovation in its autonomous driving efforts. The JD Logistics VAN features a super-large cargo capacity of 24 cubic meters, making it the highest-capacity unmanned light truck in the logistics industry, and it can replace traditional 4.2-meter trucks in logistics shuttle and station-transfer operations. Through scaled deployment and scenario optimization, vehicle cost, operational safety and efficiency are expected to be further improved.[42] This indicates that autonomous trucking is developing rapidly. Last-mile delivery (parcel/on-demand delivery) In parcel delivery, unmanned delivery vehicles have entered a stage of “large-scale procurement + batch operations.” Looking first at the demand side and sector operations, People’s Daily (October 2025) directly noted that an official from the State Post Bureau described express logistics as one of the scenarios with the “greatest potential for scaled deployment” of unmanned vehicles. The article also provided very specific figures: ZTO’s unmanned delivery fleet had exceeded 2,000 vehicles, operating in more than 200 cities, delivering more than 200,000 parcels per day, with cumulative mileage exceeding 20 million kilometers.[43] Looking next at the supply and procurement side, China Post disclosed a 2025–2026 postal district center unmanned vehicle outsourcing/leasing project. The project was divided into eight lots, with a total publicly disclosed quantity of 7,000 vehicles.[44] This shows that parcel delivery is currently one of the clearest commercialization tracks for unmanned vehicle applications. Turning to on-demand delivery/food delivery, Meituan’s financial results release states that, as of the end of 2024, its autonomous delivery vehicles and drones had cumulatively completed nearly 4.91 million and 450,000 orders, respectively, and that it is continuing to build out its autonomous delivery network.[45] Although drones are not ground AVs, from the perspective of commercial unmanned delivery, drones and unmanned delivery vehicles together reflect the pace of deployment of unmanned delivery in China’s on-demand delivery sector. In addition, Xinhua News Agency provided an industry-level snapshot in early 2025, noting that in the previous year, the express delivery industry deployed nearly 1,000 unmanned vehicles, more than 300 drones and delivered nearly 3 million parcels by drone.[46] Ports (smart terminals/unmanned transport vehicles) At a State Council Information Office press conference, it was noted that China has completed 21 automated container terminals and has 28 more under construction, and that automated guided vehicles (AGVs) and unmanned container trucks have already achieved scaled deployment.[47] As a concrete example, a 2025 Xinhua report on Tianjin Port stated that after the launch of Tianjin Port’s new-generation terminal control system, more than 100 AI horizontal transport robots had already achieved large-scale coordinated operations; meanwhile, new-energy transfer vehicles accounted for 100 percent of container terminal transfer vehicles.[48] Mining (autonomous mining trucks) At its press conference for Q4 2025, the National Mine Safety Administration stated that it would prioritize the large-scale deployment of unmanned autonomous mining trucks in open-pit mining scenarios.[49] In terms of deployment scale, by the end of 2025, Yikong Zhijia (EACON) had deployed more than 2,300 L4 unmanned mining trucks across 26 open-pit mines nationwide.[50] Agriculture In agriculture, the dominant form is, more accurately, “autonomous farm machinery (BeiDou navigation) + agricultural drones + smart farm systems.” In terms of policy and technical direction, the agriculture sector has already incorporated “unmanned operations” into its recommended key technologies. Publicly available information shows that the 2025 recommended smart agriculture technologies include “BeiDou navigation-based unmanned intelligent operation technology,” the core of which is to use high-precision positioning and multi-sensor coordinated control to enable precise autonomous operation of farm machinery.[51] XAG’s 2024 official launch materials further show that it is not only selling agricultural drones, but also promoting a full product suite including “autosteering systems for farm machinery,” “agricultural unmanned vehicles” and “intelligent irrigation and fertilization systems.” The same materials state that XAG’s channels cover 1,007 counties, its products are present in 70 countries and regions, and 466 farms are already using its smart agriculture solutions; it also explicitly notes that farm machinery equipped with its APC-series autosteering systems has already been deployed in multiple locations across China. This indicates that agricultural unmanned solutions are moving from standalone products toward ecosystem-based commercial delivery.[52]
[53] "Beijing Autonomous Vehicle Regulations" [54] "Shenzhen Special Economic Zone Intelligent Connected Vehicle Management Regulations" [55] "(Authorized to publish) Personal Information Protection Law of the People's Republic of China" [56] "(Authorized to publish) Data Security Law of the People's Republic of China" [57] "Regulations on the Management of Network Data Security (Order No. 790 of the State Council of the People's Republic of China)" [58] "The Cyberspace Administration of China and five other departments issued the Several Provisions on Automotive Data Security Management (Trial)" [59] "Notice from the Ministry of Natural Resources on Strengthening the Security Management of Surveying and Mapping Geographic Information for Intelligent Connected Vehicles" [60] "The Cyberspace Administration of China and five other departments issued the Several Provisions on Automotive Data Security Management (Trial)"
In China, autonomous vehicles (AVs) are not primarily determined by a single standalone “AI law;” rather, deployment is affected by multiple laws and regulations together. The first category is local autonomous driving regulations. These rules directly affect whether AVs can be commercialized and in which scenarios they can be deployed. For example, the Beijing Autonomous Vehicle Regulations (adopted in 2024 and effective in 2025) explicitly support the use of autonomous vehicles for private passenger cars, urban mobility services and road freight operations (excluding hazardous goods transportation). This means the scope of commercial deployment has been expressly written into local legislation.[53] Shenzhen’s earlier Regulations of the Shenzhen Special Economic Zone on Intelligent Connected Vehicles is one of the earliest systematic local regulations in China. It establishes local institutional arrangements for road access, traffic management, and the handling of traffic violations and accidents, which in practice has increased legal certainty for companies advancing L3/L4 deployment.[54] The second category is data and privacy laws. The Personal Information Protection Law (PIPL) requires personal information processing to follow the principles of legality, legitimacy, and necessity, and emphasizes data minimization; at the same time, it classifies “location and movement tracking” and “biometric information” as sensitive personal information, which requires stricter conditions and separate consent for processing. For AVs, this directly affects the design of data collection involving in-cabin and exterior cameras, DMS/OMS, facial/voice biometrics and vehicle trajectory data.[55] The Data Security Law (DSL), meanwhile, requires a data classification and tiered protection regime and a protection framework for important data, and imposes end-to-end security management obligations, risk monitoring and incident reporting requirements on data processors. These requirements directly apply to AV companies’ cloud platforms, vehicle-to-cloud data return pipelines and simulation/training data management workflows.[56] In addition, the Regulations on Network Data Security Management, effective from 2025, further refine obligations for network data processors in areas such as security safeguards, personal information notice requirements and important data management. For companies providing connected vehicle platforms and autonomous driving operations platforms, these constitute ongoing compliance requirements.[57] The third category is sector-specific automotive data rules. The Cyberspace Administration of China (CAC) and four other authorities issued the Several Provisions on the Security Management of Automotive Data (for Trial Implementation), which explicitly took effect on October 1, 2021. These provisions emphasize principles such as in-vehicle processing, no collection by default, applying appropriate precision ranges and de-identification/anonymization processing. They also require automotive data processors to fulfill personal information protection responsibilities, store important data domestically and undergo security assessments for cross-border data transfers. For autonomous driving companies, this directly affects data architecture, sensor data collection strategies and cross-border R&D collaboration models.[58] The fourth category is surveying, mapping and geospatial information regulation. In 2024, the Ministry of Natural Resources issued a notice on strengthening the security management of surveying, mapping and geoinformation for intelligent connected vehicles. The notice explicitly brings the collection and processing of spatial coordinates, images, point clouds and similar data during ICV testing/operation within the scope of surveying and mapping regulation. It also requires that navigation electronic-map-related activities be undertaken by entities with the relevant surveying and mapping qualifications, and it emphasizes domestic storage of geoinformation data, approval/review for external provision and security assessment requirements for cross-border data transfers. These rules have direct implications for HD maps, autonomous driving map updates, data return pipelines and partner qualification requirements.[59] The fifth category is the recent-detailed guidance on cross-border data transfer. On February 3, 2026, the National Data Administration and other authorities jointly issued the Automotive Data Cross-Border Security Guidelines (2026 Edition), which provide more specific rules for automotive data exports and set out several exemption scenarios from filing requirements, while still maintaining strict requirements for large-scale transfers of personal information and important data. This has significant implications for multinational automakers and cross-border R&D teams.[60]
[9] "The Robotaxi profit mystery, science fiction coming true—how far is the road to making money?" [10] "Closure, Layoffs, Delisting - 2025 Autonomous Driving "Elimination Round" Record" [11] "Xinhua Finance Survey Chinese Automakers Face Three Major Hurdles Going Global; Localization Has Become a Trend" [12] "In heavy fog and snowy weather, over 60% of driving assistance safety features fail to meet standards" [13] "2025 Auto Major Events: From Public Opinion Storm to Industry Restructuring, Regulatory Safeguards Intelligent Driving Safety First" [14] "L3-level autonomous driving pilot launches; exclusive intelligent driving insurance pending replenishment"
[29] "In 2025, 2 out of 3 new EVs in China will be equipped with Self-Driving technology" [30] "Global and China Autonomous Driving Domain Controller and CCU Report 2025" [31] "2025 Global and China Connected Autonomous Driving System Market Development In-Depth Analysis Report (IIM Information 2025 YHT65)" [32] "A panoramic and in-depth analysis of the development of China's intelligent driving industry in 2025" [33] "China’s Self-driving Tech Gains Ground on Global Roads" [34] "2025 Global and China Connected Autonomous Driving System Market Development In-Depth Analysis Report (IIM Information 2025 YHT65)" [35] "Comprehensive system in place for smart vehicles" [36] "China’s Self-driving Tech Gains Ground on Global Roads" [37] "China's Zeekr, Xpeng to offer cars with L3 autonomy as auto battle heats up" [38] "Robotaxi partnerships speed up global growth"
[15] "Understanding the Mandatory National Standard GB 44495—2024 'Technical Requirements for Vehicle Information Security'" [16] "Notice from the Ministry of Industry and Information Technology and the State Administration for Market Regulation on Further Strengthening the Management of Intelligent Connected Vehicle Product Entry, Recall, and Online Software Upgrades" [17] "The Ministry of Industry and Information Technology and eight other departments are publicly soliciting opinions on the Automotive Data Export Security Guidelines (2025 Edition)" [18] "The Ethical Guidelines for Medical Research of Neurotechnology Involving Humans and the Ethical Guidelines for R&D of Driving Automation Technology were released"
[19] https://www.news.cn/fortune/20250519/d9c4426993214d5b8410e36c6b404603/c.html [20] "China speeds up 5G-A rollout in multiple major cities," Global Times [21] "Intelligent upgrades: Driverless trucks run smoothly at Guoyuan Port" [22] "L3-level autonomous driving models receive conditional access permits, marking a critical milestone in commercial implementation"
[39] "Inceptio-Powered Autonomous Trucks Surpass 200 Million Kilometers in Commercial Operations Capping Off Groundbreaking Year" [40] https://mp.weixin.qq.com/s/5ITuNmO2yFYn3y2DPKf0nA [41] https://mp.weixin.qq.com/s/oM95Uj8ymXGtAW2AevMRuA [42] "The logistics industry has the largest cargo capacity! JD Logistics launched its self-developed VAN unmanned light truck" [43] "Unmanned Delivery Vehicles Moving Towards Scale (Big Data Observation)" [44] "China Post Express Logistics Co., Ltd. 2025 Unmanned Vehicle Rental Centralized Procurement Project (Project No.: 0747-2560SCCZAB317) Bid Candidates Announcement" [45] "Meituan's 2024 Financial Report: Full-Year Revenue of 337.6 Billion RMB, Up 22% Year-on-Year" [46] "China Express Station has reached new heights" [47] https://mp.weixin.qq.com/s/vxx1I_Ad8TEijVvm7_BzEQ；https://mp.weixin.qq.com/s/NEBL0De0h74ROCloJ4r0fQ [48] "The technological leap from unmanned docks to large models." [49] "National Mine Safety Supervision Administration Q4 regular press conference" [50] https://www.fujian.gov.cn/xwdt/fjyw/202602/t20260224_7099715.htm [51] "Ministry of Agriculture and Rural Affairs: Announcement of the Top Ten Key Smart Agriculture Technologies for 2025" [52] https://www.xag.cn/news/official/xai/1633
In 2025, autonomous vehicle (AV) technology has clearly transitioned from experimental prototypes toward early commercial deployment, large-scale pilots and increasingly standardized platforms. Progress is visible across passenger mobility, logistics, software-defined vehicles, connectivity and regulatory frameworks, indicating that autonomy is moving into structured real-world applications rather than remaining a purely experimental technology. One of the most visible developments has been the expansion of Level 3 to Level 4 autonomous driving systems into consumer-facing and commercial services. A popular electric vehicle company initiated its robotaxi service in Austin, Texas, operating without a safety driver, marking one of the first fully autonomous ride-hailing offerings accessible to the public. Waymo has further expanded its robotaxi operations in Phoenix and San Francisco and is preparing entry into additional global cities, including London and Tokyo. In parallel, Waymo’s partnership with Uber to roll out driverless taxis in cities such as Atlanta highlights how autonomy is being integrated directly into existing ride-hailing ecosystems. These developments demonstrate a decisive shift from closed testing environments toward real urban traffic and commercial operations. In Europe, Stellantis and Bolt announced plans for scalable Level 4 driverless ride-hailing trials, starting in 2026, representing the first coordinated European effort to deploy passenger robotaxis at scale. German OEMs such as Mercedes-Benz, BMW and Volkswagen continue to focus on Level 2+ and early Level 3 systems, with Mercedes’ DRIVE PILOT receiving regulatory approval in Germany and Nevada. Meanwhile, China approved its first batch of Level 3 autonomous vehicles, enabling hands-free driving under defined operational conditions and signaling growing regulatory acceptance of conditional automation for private consumers. Chinese AV developers such as Pony.ai have also begun mass production of seventh-generation robotaxi platforms, targeting fleets of more than 1,000 Level 4 vehicles in cities including Beijing, Shenzhen and Wuhan by the end of 2025. Technological advances underpinning these deployments are equally significant. Advanced AI algorithms have substantially improved object detection, decision-making and route optimization. Companies are increasingly experimenting with generative AI to support real-time data processing, predictive analytics and adaptive behavior in complex traffic scenarios. A key example is NVIDIA’s DRIVE Thor system-on-a-chip, which offers roughly 20 times the computing power of its predecessor and integrates data from cameras, radar and LiDAR to support real-time autonomous navigation. This level of computing enables a single platform to serve both passenger vehicles and broader robotics applications. Sensor technology continues to evolve as well. High-resolution cameras, radar and next-generation LiDAR systems are becoming more affordable and capable, supporting Level 3 and higher automation. Startups such as South Korea–based AutoL are developing long-range LiDAR sensors capable of detecting objects at distances of up to 300 meters, significantly enhancing perception in both highway and urban environments. At the same time, extended reality technologies are playing a growing role: virtual reality is used to simulate complex driving scenarios for machine learning and validation, while augmented reality supports advanced head-up displays and new customer-facing experiences such as virtual showrooms. Connectivity has become a critical enabler of autonomy in 2025. Vehicle-to-everything (V2X) communication is being rolled out in many cities, allowing vehicles to exchange real-time data with infrastructure, traffic systems and other vehicles. Combined with IoT-enabled sensors, this connectivity improves situational awareness and supports safer, more efficient decision-making. Looking ahead, 5G Advanced promises even lower latency and higher reliability for connected vehicles, while early research into 6G points toward ultra-reliable, intelligent communication at scale – an essential prerequisite for fully autonomous transportation systems. Commercial logistics represents another major growth area. Companies such as Aurora and TuSimple are preparing to deploy hundreds of autonomous trucks on major freight corridors by 2026, following extensive testing in the United States and China. These systems aim to reduce costs, address driver shortages and improve safety in long-haul transport. In contrast, intra-city autonomous delivery remains more challenging due to tasks such as loading, unloading and customer interaction. Nevertheless, companies such as Nuro are already operating compact, electric autonomous delivery pods in select U.S. cities, delivering groceries and prepared meals without human drivers. Large retailers and logistics players, including Amazon, Walmart and FedEx, continue to invest heavily in this space. A notable structural trend in 2025 is the rise of software-driven collaboration and open-source platforms. Software-defined vehicle (SDV) architectures enable modular design, continuous over-the-air updates and scalable autonomy across vehicle platforms. The S-CORE Project, announced in June 2025 and backed by companies such as Bosch, QNX and Mercedes-Benz, aims to establish the first open-source core stack for SDVs, standardizing middleware layers with a strong focus on functional safety. Such initiatives are intended to reduce fragmentation and accelerate large-scale AV deployment. With increasing connectivity and automation, cybersecurity has become a central concern. Autonomous vehicles are exposed to risks ranging from data breaches and ransomware to vehicle hijacking and GPS spoofing. As a result, multi-layered security architectures, including end-to-end encryption, intrusion detection systems and AI-driven threat monitoring, are being implemented. Regulators are also moving toward stricter cybersecurity standards, while OTA updates allow manufacturers to deploy security patches rapidly without physical recalls. In parallel, blockchain technologies are being explored to enable tamper-proof data management, automated insurance claims, toll payments and even data monetization models for vehicle owners. Finally, 2025 has seen autonomous driving increasingly embedded within broader Mobility-as-a-Service (MaaS) concepts. robotaxis, autonomous shuttles and shared AV fleets are being integrated into urban mobility ecosystems. Overall, the latest developments in 2025 reflect a convergence of technological maturity, regulatory progress and new business models. Autonomous vehicles are no longer confined to test tracks or limited pilots. Instead, they are entering structured commercial use across passenger mobility and logistics. With continued advances in AI, connectivity, cybersecurity and software platforms, the industry is laying the groundwork for broader adoption in 2026 and beyond, potentially reshaping ride-hailing, freight transport and urban mobility as a whole.
Despite progress, multiple challenges remain. Germany – and Europe more broadly – is currently experiencing several significant roadblocks that are slowing the development and large-scale deployment of autonomous vehicles, despite clear political support and technological capability. At the European level, regulatory complexity remains a key barrier. Harmonized frameworks such as UNECE (United Nations Economic Commission for Europe) Regulation No. 157 and the EU General Safety Regulation II impose stringent type-approval, safety validation and monitoring requirements. While these rules strengthen safety and public trust, they also slow the rollout of Level 3 and Level 4 systems. Autonomous driving projects are still largely confined to small, isolated pilot programs (for example in Hamburg) with little cross-border alignment, making scalable, pan-European deployment difficult. Strict data protection and cybersecurity requirements under the GDPR (General Data Protection Regulation) further complicate the use of sensor and video data, and there is still no fully transparent, standardized incident reporting regime across Europe. From a strategic perspective, Europe faces a broader challenge: despite political calls for an “AI-first” approach and homegrown autonomous technology, the market is increasingly opening to non-European players, particularly Chinese AV companies partnering with global mobility platforms. This raises concerns that Europe may become more of a consumer market for autonomous technologies than a global developer and exporter. In Germany, the legal framework formally allows Level 4 autonomous vehicles in defined operational domains, but recent legislative updates have added further complexity, especially regarding cybersecurity certification, data governance and detailed handover rules between humans and automated systems. While these measures improve legal certainty, they significantly increase compliance effort and slow deployment. Market conditions also pose challenges: high private car ownership, a strong taxi lobby and a well-developed public transport system leave limited room for new autonomous mobility services. Many pilot projects have struggled to continue once public funding ended. Finally, German manufacturers remain highly cautious. According to the ADAC (Allgemeiner Deutscher Automobil-Club), OEMs are reluctant to enter series production with systems they consider insufficiently mature. Although city trials have so far been accident-free, AVs’ overly defensive driving behavior has sometimes caused frustration among other road users, highlighting issues of acceptance in mixed traffic. Overall, progress in Germany and Europe is steady but slow, constrained by regulatory rigor, fragmented pilots and unresolved questions around scalability and long-term deployment.
Yes. Cybersecurity and privacy have become core requirements for the safe deployment of autonomous vehicles, and recent regulatory and technical updates reflect this shift. Autonomous vehicles depend on connected sensors, V2X communication, cloud-based AI and over-the-air (OTA) software updates, which significantly increases their exposure to cyber risks. In Germany and across Europe, OEMs must comply with national road traffic laws, the GDPR and the UNECE cybersecurity and software update regulations (R155 and R156). Recent updates in 2025 place stronger emphasis on continuous cybersecurity validation, post-market monitoring, secure OTA updates, encryption, intrusion detection and mandatory reporting of safety-critical software changes. Vehicle software is now treated as a continuously evolving system rather than as a static product. These rules address a growing threat landscape. Autonomous vehicles integrate many interconnected subsystems, meaning that vulnerabilities in one component can affect safety-critical functions. AVs are particularly vulnerable to sensor attacks such as spoofing or jamming of cameras, radar and LiDAR, as well as data breaches involving sensitive location and passenger data. Denial-of-service attacks targeting communication or computing resources also pose serious operational risks. To mitigate these threats, manufacturers rely on layered security measures, including intrusion detection systems, strong encryption, authentication protocols and regular OTA updates to close vulnerabilities. Cybersecurity management systems and formal risk assessments are now standard throughout the vehicle lifecycle. At the same time, privacy regulation is evolving alongside cybersecurity. Autonomous vehicles collect large amounts of sensor and location data, which must be minimized, anonymized and processed in line with data protection laws such as the GDPR. Regulators increasingly require transparency and explainability, including system logs that allow reconstruction of vehicle decisions for accident investigations and liability assessment. Overall, recent updates show a clear move toward continuous, lifecycle-based cybersecurity and privacy compliance. These measures are no longer optional add-ons, but essential foundations for public trust and scalable autonomous vehicle deployment. The risks described above are not merely theoretical. In 2024, a significant data issue exposed approximately 800,000 EV owner datasets due to a cloud storage misconfiguration. The incident illustrates that the primary cybersecurity risk for AV and EV fleets in the near term is not dramatic remote vehicle takeover, but rather data exposure arising from inadequately secured cloud and backend infrastructure. It also demonstrates the multi-jurisdictional litigation risk.
5G continues to play a critical role in enabling autonomous driving, particularly through ultra-low-latency and high-reliability vehicle-to-everything (V2X) communication. 5G networks support real-time interaction between vehicles, road infrastructure and cloud-based traffic management systems, allowing autonomous vehicles to safely navigate complex and dynamic urban environments. When combined with edge computing, 5G enables fast, local processing of large volumes of sensor data, significantly improving reaction times, predictive path planning, collision avoidance and fleet coordination. Automaker–telecom collaboration has intensified as autonomy scales. Partnerships such as Hyundai and Avride illustrate how Level 2+ to Level 4 robotaxis increasingly rely on 5G (and future 6G-ready) edge networks to support automated lane changes, cooperative driving, remote monitoring and continuous over-the-air software updates. In logistics and industrial use cases, 5G has become a backbone technology for autonomous trucks and delivery robots, enabling reliable operation in dense urban areas and supporting large-scale fleet orchestration. A key industry player in this space is Ericsson, which has emerged as a leading provider of 5G infrastructure tailored to automotive requirements. Ericsson works closely with automakers to integrate 5G into connected and autonomous vehicles, with transport and traffic authorities to accelerate smart transportation systems, and with technology providers to build end-to-end 5G and future 6G ecosystems. In parallel, the company is actively involved in global standards development through organizations such as 3GPP, ITU, ETSI and industry alliances such as the 5G Automotive Association (5GAA), helping to harmonize technical requirements for V2X and autonomous driving worldwide. Looking ahead, several trends are emerging alongside 5G deployment: collaborative AI, where vehicles exchange data to improve collective decision-making; global expansion of autonomous fleets into new markets; and a growing focus on sustainability, including the integration of renewable energy into connected mobility systems. For businesses, adapting to these developments means investing in R&D, building strong partnerships across the automotive and telecom sectors, and upskilling the workforce in AI, 5G and autonomous systems. Overall, recent updates confirm that 5G is not just an enabling technology, but a foundational pillar for scalable, safe and commercially viable autonomous vehicle deployment.
Yes, several critical regulatory updates occurred globally in 2025: In Germany, the Autonomous Driving Act (2021) continues to allow Level 4 vehicles in defined operational domains, with clear allocation of liability between manufacturers, operators and insurers. Recent updates have refined this framework by strengthening cybersecurity, data protection and driver handover requirements. The Autonomous Vehicles Approval and Operation Ordinance (AFGBV) has clarified approval and operational procedures for Level 4 vehicles, while the Road Traffic Remote Control Regulation (StVFernLV), in force since December 2025, introduces a legal basis for remotely controlled vehicles, opening new use cases in car sharing, taxis and logistics. The StVFernLV distinguishes remotely controlled vehicles — operated by a human operator at a distance — from autonomous vehicles acting on their own AI-based decisions. This distinction is legally significant: liability for a remotely controlled vehicle will generally rest on the human remote operator, whereas liability for a genuinely autonomous decision rests on the ADS and, under the new EU Product Liability Directive, on the manufacturer. At the EU and UNECE level, the General Safety Regulation II (GSR II) mandates advanced driver assistance systems in new vehicles, and type-approval rules were expanded in 2025 to cover Automated Valet Parking (AVP) systems. Ongoing amendments to UNECE Regulation No. 157 (ALKS) and the introduction of UNECE Regulation No. 171 (DCAS) are particularly impactful. These updates raise operational limits (e.g., higher speeds and automated lane changes), strengthen driver monitoring and cybersecurity requirements, and clarify system behavior. While they improve legal certainty, they also explain why many OEMs remain cautious and continue to market systems as Level 2+ rather than full Level 3. At UNECE level, the Working Party on Automated/Autonomous and Connected Vehicles (GRVA) is advancing a proposal for a new Global Technical Regulation (GTR) on Automated Driving Systems. Unlike existing regulations that govern specific functions (ALKS, DCAS, AVP), the proposed ADS GTR would establish a comprehensive global standard for the entire automated driving system. If adopted, it would provide a harmonized benchmark for safety validation, type approval, and standard-of-care analysis in product liability proceedings worldwide. A notable political development is the Joint Declaration of Intent on EU Cross-Border Testbeds for the Deployment of Automated Vehicles, signed by multiple Member States, aiming to establish harmonized cross-border test corridors. A further development at EU level is the new Product Liability Directive (Directive 2024/2853). For the first time, software and AI are explicitly classified as 'products' capable of giving rise to product liability claims. The Directive introduces statutory presumptions of defect and causation, significantly lowering the evidentiary burden for injured parties in AV-related claims. This development is of direct relevance to manufacturers, fleet operators, and defense counsel: the established principle that plaintiffs must fully prove defect and causation no longer applies without qualification where software or AI systems are involved. Compliance documentation, safety cases, and ODD validation records will become primary litigation tools — either as evidence of due care or, in the hands of plaintiffs, as proof of known limitations. In China, regulation has accelerated through national pilots and local laws. Beijing’s Autonomous Vehicle Regulation, effective April 2025, establishes a comprehensive framework for Level 3–5 vehicles, including licensing, liability, safety and data obligations. Additional rules now require regulatory approval for safety-relevant OTA updates and stricter driver monitoring for Level 2+ systems. In the United States, progress remains fragmented. NHTSA expanded exemptions from federal safety standards, updated crash reporting and proposed the AV STEP Rule to improve deployment transparency, but no comprehensive federal AV law exists. As a result, individual states continue to lead, with widely differing rules on testing, deployment, insurance and operator responsibility. These regulatory changes collectively reflect a global trend toward structured, regulated commercialization of autonomous vehicles, moving beyond pilot testing to defined operational frameworks with clear accountability.
In 2026, Germany and other EU countries are expected to make tangible progress toward scalable and commercially viable autonomous vehicle deployment, moving beyond isolated pilots toward more structured operations under mature regulatory frameworks. In passenger mobility, the most visible advancement will likely be the expansion of Level 4 robotaxi services in defined operational design domains (ODDs). Germany is expected to remain a frontrunner within Europe, with cities such as Munich and Hamburg continuing to act as regulatory sandboxes for autonomous mobility. Public pilot services may transition into limited commercial offerings, supported by Germany’s Autonomous Driving Act and detailed implementation rules under the AFGBV. A key signal in this direction is the announcement by Uber and Momenta, which plan to test Level 4 autonomous vehicles in Germany starting in 2026, integrating Momenta’s AI-based driving technology into Uber’s platform with the goal of scaling robotaxi services across Europe. In consumer vehicles, 2026 should see a broader rollout of Level 3 conditional automation systems in Germany and select EU markets. Systems such as Mercedes-Benz’s DRIVE PILOT are expected to expand in availability and functionality, supported by over-the-air updates, improved sensor fusion and more capable AI software. European OEMs will continue to take a cautious, compliance-driven approach, but conditional automation on highways is likely to become more common in premium vehicle segments. In commercial logistics, autonomous driving is expected to scale more quickly. Germany and neighboring EU countries are likely to see increased use of autonomous or highly automated trucks on highways, including platooning concepts, as well as pilot deployments of autonomous last-mile delivery vehicles in urban areas. These use cases benefit from more predictable routes and clearer business cases, making them attractive early adopters within the EU regulatory environment. From a technology and infrastructure perspective, 2026 will bring deeper integration of 5G and satellite connectivity across Europe. German OEMs such as BMW are actively leveraging 5G and satellite networks to enhance vehicle connectivity, enable new data-driven services and potentially create new revenue streams by turning vehicles into mobile data platforms. At the same time, wider deployment of 5G and edge computing across the EU will support low-latency vehicle-to-infrastructure (V2I) communication, fleet coordination and remote monitoring of autonomous systems. On the regulatory side, continued alignment at EU and UNECE level (particularly around ALKS, DCAS and Automated Valet Parking) will further facilitate cross-border operation of automated vehicles. While regulatory rigor will remain high, this harmonization is expected to reduce uncertainty and make multi-country deployment within the EU more feasible. Overall, 2026 is likely to mark a clear transition in Germany and the EU from experimentation toward operational deployment at limited scale, especially in robotaxis, highway automation and logistics. Progress will remain cautious and compliance-focused, but the foundations for a sustainable European autonomous mobility ecosystem will be firmly in place.
Yes. Beyond robotaxis, autonomous vehicle technology is advancing rapidly across a wide range of commercial and industrial applications, where the business case is often clearer and deployment conditions are more controlled. In freight and logistics, autonomous driving is progressing steadily toward commercial use. Driverless or highly automated trucks are being deployed on highways and fixed freight corridors, supported by fleet orchestration platforms, advanced sensor fusion and AI-based remote monitoring. These systems aim to address driver shortages, improve safety and reduce operating costs. In parallel, autonomous urban delivery fleets and warehouse automation are expanding, with robots and drones increasingly used for parcel handling, inventory monitoring and last-mile logistics. Public transport and shuttle services are another important growth area. Level 4 autonomous buses and shuttles are entering pilot and early operational programs in business parks, campuses, airports and industrial zones. These environments offer predictable routes and controlled traffic conditions, making them well suited for early deployment of driverless public transport solutions. A particularly strong area of development is agriculture and industrial vehicles. Autonomous farming machinery is already market-ready in several use cases, especially in Germany and other parts of Europe. Field robots and (semi-)autonomous tractors are used for sowing, weeding and crop treatment, helping farmers cope with labor shortages and reduce physical strain. Using GPS, cameras, sensors and AI-based image recognition, these machines can apply fertilizer and crop protection products with high precision, increasing yields while reducing input use. Lightweight field robots also help reduce soil compaction. Autonomous weeding systems, especially in vegetable and organic farming, can operate 24/7 at low speeds and have demonstrated time savings of up to 80 percent compared to manual labor. Major manufacturers such as Claas, John Deere, Case and Fendt have developed autonomous tractors, and Fraunhofer IOSB introduced a retrofit kit allowing tractors to be remotely operated over 200 km. Public funding programs in German states support adoption, particularly for organic or labor-intensive crops. In delivery and food logistics, ground-based delivery robots, autonomous vehicles and drones are increasingly used for last-mile delivery. Examples include: Germany: Grocery retailer Rewe is piloting a Level 4 autonomous VW ID. Buzz for delivering groceries in Bochum, in collaboration with Swiss startup Loxo. Europe: At the European University in Villaviciosa de Odón (Spain), Ottobots from Ottonomy autonomously deliver food and drinks across campus in partnership with Sodexo. Global warehouse and food logistics: Robotic arms are widely deployed in a leading global technology and e-commerce company fulfillment centers for package handling, and food-tech applications use automation for fry stations, inventory management and quality control.
Yes. AI regulations are becoming increasingly important for the deployment of autonomous vehicles in Germany and across the EU. The EU AI Act (Regulation (EU) 2024/1689), set to be widely enforced from August 2026, classifies most AI-driven autonomous driving and assistance systems as high-risk AI (HRAI) due to their critical role in on-road decision-making. Key requirements for AV deployment under the AI Act include: Explainable AI and decision logging: AV systems must maintain detailed records of decisions to allow post-incident reconstruction and support transparency and accountability. Data governance (Article 10): Training, validation and testing data must be accurate, representative and free from bias, covering all relevant scenarios, including edge cases. Simulation and rigorous annotation are required to ensure safe and reliable operation. Human oversight (Article 14): High-risk AI systems must be designed to allow effective supervision by humans during use. Accuracy, robustness and cybersecurity (Article 15): AV AI systems must be resilient against faults and cyberattacks, performing reliably throughout their lifecycle. While sector-specific automotive regulations (e.g., UNECE standards, Germany’s Autonomous Driving Act) remain the primary regulatory framework, the AI Act serves as a supplementary layer, providing harmonized standards for safety, transparency, accountability and cybersecurity. Providers often perform self-assessment of high-risk classification, documenting why a system qualifies or not, which accelerates market entry, but requires rigorous post-market monitoring. Overall, these AI requirements ensure that autonomous vehicles operate safely, transparently and in compliance with EU standards, influencing software development, OTA updates, human-machine interaction protocols and liability frameworks. They mark a significant step toward standardized, high-assurance AV deployment in Germany and across Europe. Alongside the EU AI Act, the new Product Liability Directive 2024/2853 further transforms the legal landscape for AI-driven AV systems. By classifying AI as a 'product' and introducing presumptions of defect and causation, the Directive creates a direct legal bridge between AI Act compliance obligations — particularly data governance, explainability, and human oversight— and civil liability exposure. Non-compliance with AI Act requirements will therefore not only carry regulatory consequences but will also constitute material evidence of defect in product liability proceedings under the new Directive.
While India continues to be at a nascent stage in the adoption and deployment of autonomous vehicles (AVs), there has been sustained momentum in the Indian automotive ecosystem, particularly in integrating emerging technologies such as artificial intelligence (AI), Internet of Things (IoT) and 5G connectivity into vehicular systems. Although the legislative framework remains limited, industry-led efforts, academic collaborations and public-private pilot initiatives are gradually shaping India’s AV trajectory. In a significant development, India unveiled its first fully indigenous AV prototype, developed under the Wipro-IISc Research and Innovation Network (WIRIN). The prototype, which was showcased in Bengaluru in October 2025, is equipped with AI, robotics, computer vision and 5G-enabled communication, and has been tested in controlled campus environments to demonstrate driverless navigation in complex Indian traffic conditions. The vehicle remains under testing, with further efforts being made to expand real-world mapping coverage. Similarly, Minus Zero, a Bengaluru-based start-up, unveiled a proprietary AI-powered vision-based autopilot system aimed at unstructured urban traffic. The system, which does not rely on traditional LiDAR or HD mapping, is trained using large-scale self-supervised learning techniques and is being tested on Indian roads under human supervision. Another notable advancement was the launch of “Swayamgati,” India’s first production-ready autonomous electric three-wheeler, by Omega Seiki Mobility. The vehicle, designed for operation in controlled environments such as airports, campuses and industrial parks, is equipped with LiDAR, GPS and AI-based obstacle detection technology, and offers a range of up to 120 km on a single charge. It is expected to be deployed on pre-mapped routes under structured commercial arrangements. In parallel, industry focus on enhancing advanced driver-assistance systems (ADAS) has deepened. Indian engineering teams are exploring the application of ultra-black coatings within LiDAR and optical sensor housings to suppress stray light and glare, improving the reliability and safety of sensors under harsh lighting conditions. These developments suggest growing industry maturity and innovation in foundational AV technologies.
India’s progression toward adopting AVs continues to face structural, regulatory and policy-related roadblocks. While technological development remains active through both public and private initiatives, deployment of fully autonomous systems on Indian roads remains constrained. One of the principal challenges is the lack of a comprehensive legal and regulatory framework to govern AV operations. Current statutes such as the Motor Vehicles Act, 1988 (MV Act) and associated rules do not provide for a driverless vehicle operating without human intervention. Key aspects such as fault attribution, testing protocols, liability, insurance and operational standards for AVs remain unaddressed under existing legislation. The complexity of India’s road infrastructure further compounds this challenge. Unpredictable traffic patterns, poor lane discipline, mixed vehicular presence (including pedestrians, livestock and non-motorized vehicles) and inconsistent signage and road markings make it difficult for advanced AV systems to operate reliably without extensive localization and adaptation. Policymakers have also expressed reservations regarding the impact of AVs on employment. The Union Minister for Road Transport and Highways has publicly reiterated his apprehension about driverless vehicles in India, citing concerns over potential job losses among professional drivers. These statements reflect a protectionist sentiment that continues to shape the government stance toward AV deployment. Furthermore, public infrastructure, including digital mapping, real-time traffic data and connectivity layers required for AV deployment, is not yet fully developed or standardized across jurisdictions. In the absence of statutory guidance, industry players remain hesitant to conduct real-world AV trials beyond limited, closed environments. As a result, AV-related activities in India remain largely confined to controlled pilot projects or academic and R&D programs. Broader public road deployment will likely depend on legislative reform, investment in digital infrastructure and alignment across various stakeholders.
AVs rely heavily on data-intensive systems, including location data, in-cabin imaging and behavioral inputs. This gives rise to heightened privacy and cybersecurity concerns, especially as vehicle connectivity becomes more advanced. Recognizing these concerns, the Indian regulatory framework has seen notable developments in 2025. The most significant development is the notification of the Digital Personal Data Protection Rules, 2025 (DPDP Rules) under the Digital Personal Data Protection Act, 2023 (DPDP Act). The DPDP Rules, issued in November 2025, lay out specific compliance requirements for organizations handling personal data in India, including data minimization, purpose limitation, data breach reporting and cross-border data transfer restrictions. A phased implementation roadmap has been proposed for the DPDP Rules, with an extended transition window of up to 18 months to enable readiness. Given that AVs often collect and process sensitive data such as biometric identifiers, precise location history and video imaging of occupants and external surroundings, the DPDP framework is expected to have a material impact on how AV developers and operators manage data. Entities engaged in AV research or deployment will need to build systems that incorporate consent-based data collection, implement robust encryption and maintain data residency where applicable. In addition, there have been efforts to strengthen automotive cybersecurity through collaborative initiatives. The Indian Institute of Technology, Kanpur and the Automotive Research Association of India (ARAI) entered into a memorandum of understanding to jointly develop cyber-resilient technologies for the automotive sector. This initiative aims to address emerging threats to connected and autonomous mobility systems by creating robust cybersecurity frameworks and conducting joint research. Meanwhile, the ARAI continues to work on finalizing Automotive Industry Standards on security and privacy for connected vehicles. Although still in draft form, these standards are expected to include technical requirements for data anonymization, access control and cyber resilience.
Connectivity plays a central role in enabling vehicle-to-everything communication, which is essential for real-time decision-making in AVs. In India, 5G deployment continues to expand, and policy discourse around machine-to-machine (M2M) connectivity has seen significant engagement from regulators and industry in 2025. In April 2025, the Telecom Regulatory Authority of India (“TRAI”) released its final recommendations on issues relating to critical M2M and IoT services, following a 2024 consultation. These recommendations aim to establish a framework to govern ultra-reliable, low-latency communication services, which are a foundational requirement for AV systems. TRAI proposed a sectoral consultation process between the Department of Telecommunications under the government of India and relevant sectoral regulators to designate “critical” M2M services. Importantly, TRAI’s final view did not mandate exclusive use of licensed spectrum for such services and instead, it recommended a performance-based approach that allows for the use of both licensed (e.g., 5G) and unlicensed (e.g., Wi-Fi 6) spectrum, provided that minimum benchmarks for latency, availability and quality are met. This flexible model could allow AV developers to choose from a range of connectivity options to suit deployment scenarios, particularly in semi-structured environments such as industrial zones, ports or educational institutions. Further, telecom service providers such as Vodafone Idea have in their recent consultation responses urged for a strong regulatory architecture for AVs, including proposals for strict liability regimes and mandatory human oversight for high-risk automated decisions. However, such recommendations have not yet been codified into enforceable regulation. While there is no AV-specific spectrum policy or deployment mandate in place, India’s evolving M2M framework and 5G rollout continue to enhance the enabling environment for AV connectivity, particularly in controlled use cases.
As of January 2026, India has not enacted any dedicated legislation or regulatory framework specific to the deployment or operation of AVs. The operation of vehicles on Indian roads continues to be governed by the MV Act and the rules framed thereunder, alongside sectoral laws such as the Consumer Protection Act, 2019 (CPA). However, neither the MV Act nor the CPA explicitly address issues unique to AVs, such as driverless operation, system-level fault attribution or testing protocols. While a legislative gap persists, the regulatory landscape has seen peripheral activity that may indirectly influence AV development. In particular, 2025 witnessed a regulatory push toward mandating ADAS, which may serve as a stepping-stone to broader automation. Notably, the Ministry of Road Transport and Highways under the Government of India (MoRTH) issued a draft notification mandating the inclusion of key ADAS features, such as automatic emergency braking systems, lane departure warning systems and driver drowsiness attention warning systems, in new passenger transport vehicles with more than eight seats starting in April 2026. In a parallel development, MoRTH approved regulations requiring the inclusion of these features in all new trucks and buses by October 2027, with compliance for existing models phased in by January 2028. Although these developments stop short of regulating AVs directly, they demonstrate a growing emphasis on integrating intelligent safety systems into vehicles, potentially laying the groundwork for future regulatory acceptance of higher levels of automation. Industry commentators and safety advocates have reiterated the need for a comprehensive AV regulatory framework. This highlights the need for standard operating procedures covering aspects such as road testing, liability, insurance and cybersecurity. However, no formal law or draft bill on AVs has been introduced for public consultation thus far. In summary, while 2025 saw no legislative advances specific to AVs, developments around ADAS mandates and policy dialog suggest a gradual move toward enabling a regulatory environment that could accommodate future AV technologies.
The AV landscape in India is expected to continue evolving in 2026, with advancements anticipated primarily in semi-autonomous systems, controlled-environment pilots and policy readiness. While full autonomy remains a long-term goal, industry data and public-private initiatives suggest incremental progress is already underway. A key trend is the expanding adoption of ADAS, particularly in the passenger vehicle segment. According to market data, the share of new vehicles equipped with ADAS features grew from approximately 6.2 percent in the first half of 2024 to 8.3 percent in the first half of 2025, showcasing a 33 percent year-on-year increase. Features such as adaptive cruise control, automated emergency braking and lane-keeping assistance are expected to become more prevalent in mid-range vehicle models. In parallel, Indian original equipment manufacturers and startups are expected to scale up pilot deployments of AV technologies tailored to India’s challenging traffic environments. These may include low-speed, geo-fenced operations in campuses, logistics parks and industrial areas, as evidenced by the recent trials of Omega Seiki’s autonomous electric three-wheeler and the WIRIN prototype discussed above. Further, companies such as Minus Zero are expected to advance AI-based urban autopilot systems trained for Indian traffic scenarios. While these systems remain in validation phases and are operated under human supervision, their development indicates potential future integration into commercial fleets or last-mile logistics platforms. On the policy front, implementation of the DPDP Rules and evolving M2M connectivity standards will likely shape how AV systems manage data, connectivity and compliance obligations. Overall, 2026 is expected to witness broader deployment of ADAS Level 2/2+ features, greater experimentation with semi-autonomous platforms in structured settings and gradual improvements in the legal and technical ecosystem. The path to ADAS Level 4/5 autonomy remains aspirational, but is being paved by a combination of regulatory nudges and industry-led adaptation to Indian conditions.
India’s AV activity beyond passenger mobility continues to evolve across multiple sectors, including logistics, agriculture and industrial automation. As discussed above, commercial AV deployments are being tested and adopted in structured or semi-structured environments where regulatory uncertainty and infrastructural challenges can be better managed. In 2025, Tata Motors filed for a patent for its “Tata YU” concept vehicle, which is an autonomous, electric, multimodal platform envisioned for last-mile goods and passenger transport. Though still at a conceptual stage, the Tata YU reflects OEM-level interest in dual-purpose autonomous solutions for urban and intra-campus logistics. Indian startups such as Flux Auto introduced “Odin,” a software platform enabling ADAS Level 4 autonomy in heavy industrial vehicles. Odin supports applications in warehouses, ports and mining environments, automating functions such as goods movement, loading/unloading and site navigation. The solution is designed to retrofit existing fleets and has been deployed in select industrial facilities for early trials. In the logistics and hospitality sectors, Autofina Robotics introduced the “PuduBot 2” delivery robot in India. The platform features 360-degree perception, dual LiDAR and autonomous mapping capabilities, and has been deployed in hospitals, restaurants and corporate offices in metropolitan centers for indoor deliveries. Agriculture is another area of interest with Indian startups such as Ati Motors having deployed the “Sherpa” autonomous mobile robot for payload transport in factories and farms. These units navigate autonomously across mixed terrain and are capable of supporting automation in harvesting and materials movement. Such developments indicate a growing trend of AV use cases emerging outside public-road passenger transport. Structured commercial environments with defined parameters remain the most conducive to early AV adoption in India, and stakeholders are increasingly leveraging autonomy in these domains to improve efficiency and reduce human intervention.
As of January 2026, there are no binding laws in India that specifically regulate AI systems or their deployment in AVs. While AV systems rely heavily on AI for perception, decision-making and control, the current legislative and regulatory environment in India continues to treat AI applications under general technology and sectoral laws. In 2025, the Ministry of Electronics and Information Technology under the Government of India issued the IndiaAI Mission’s National AI Governance Guidelines. These guidelines aim to promote responsible AI development and adoption across sectors, and outline voluntary principles such as fairness, safety, transparency, accountability and human oversight for high-risk AI use cases. However, these guidelines are not legally binding and do not contain AV-specific directives. They also do not prescribe rules for testing, certification or liability in relation to AI-driven AV systems. While the principles may indirectly influence AV design and deployment, particularly with regard to explainability and safety, they do not create new obligations under law. It is also pertinent to note that AVs will be indirectly impacted by the implementation of the DPDP Act and DPDP Rules, to the extent that such systems process personal data. However, this impact arises from data protection legislation rather than from any AI-specific legal regime. There have been stakeholder discussions around the need to regulate autonomous systems, including through TRAI’s consultations on critical M2M services; however, no draft legislation has been published to date. Accordingly, the deployment of AI in AVs in India continues to operate without a bespoke legal framework, although regulatory expectations around safety, data protection and accountability are beginning to evolve.
New temporary autonomous vehicle operation permits decreased in 2024-2025 The number of applications for new “temporary autonomous vehicle operation permits,” which allow developed autonomous vehicles to be tested on actual roads without official registration, saw increases for four years, but has shown a downward trend since 2024 due to the sluggish electric vehicle industry, including electric autonomous vehicles. Specifically, there were 43 new applications in 2020, 66 in 2021, 86 in 2022, 151 in 2023, 41 in 2024 and a slight rise in the number to 44 as of September 2025. Hyundai Mobis's Level 4 autonomous vehicle runs on the roads in Incheon Hyundai Mobis, an affiliate of the Hyundai Motor Group, conducted test drives of its Level 4 autonomous vehicle prototype around the Songdo and Yeongjongdo areas in Incheon. To run a test vehicle, Hyundai Mobis signed a Memorandum of Understanding (MOU) with the Incheon Free Economic Zone Authority and Incheon Technopark.[1] The autonomous vehicle prototypes were scheduled to operate across a 60-km section during the first half of 2025. Through this pilot program, Hyundai Mobis aims to validate the capabilities of its self-developed autonomous driving solutions. The autonomous vehicles are equipped with computing technologies optimized for Level 4 autonomy and designed to collect meaningful data through extended urban driving. Also, Hyundai Mobis plans to concurrently develop mass-production-level systems through this pilot program. To achieve this, the prototype vehicles are equipped with high-performance processors that replace dozens of microcontroller units (MCUs) responsible for vehicle control. Autonomous public transportation The Ministry of Land, Infrastructure and Transport (MOLIT) and the Korea Transportation Safety Authority (TS) implemented an automated driving shuttle bus system around Gyeongju during the APEC event that took place from late October to November 2025. The self-driving shuttles, “K-Autonomous,” combined hardware from leading Korean automakers, providing those automakers with an opportunity to showcase Korea’s homegrown autonomous driving technology on the global stage. Trial operations of self-driving freight trucks can now be conducted on all highways in Korea, which could lead to the start of long-haul autonomous commercial logistics services nationwide in the future. The Seoul Metropolitan Government announced that, after about eight months of pilot operation, the service area of KG Mobility vehicles expanded to all of Gangnam starting on June 16, 2025. The pilot program continues with three autonomous taxis. Seoul reported roughly 4,200 rides with zero accidents during the initial period.
[1] Further details available at https://www.iotworldtoday.com/transportation-logistics/hyundai-mobis-plans-major-self-driving-test-in-south-korea#close-modal.
South Korea aims to commercialize Level 4 in 2027, but the field is criticized for the serious technological and structural gap compared to global leaders such as Waymo in the U.S. that have already become a means of daily transportation, with flexible and smooth driving that allows them to reverse themselves for vehicles coming from narrow intersections. South Korea is still hampered by early-stage regulations at the level of a decade ago, failing to properly conduct driverless tests. In particular, regulations that force safety personnel to board seats are hindering the essential development of autonomous driving technologies, and companies are hiding behind these regulatory barriers and showing a passive attitude toward technological innovation and investment. Therefore, if South Korea is not to be subordinated to foreign technologies such as the global OS or search engine market in 10 years, it should completely lift the bureaucracy in technology and industry, and significantly flex the regulations to allow driverless driving in front of schools or senior protection zones, depending on their performance level. In addition, it is urgent to transparently disclose the technology level of each manufacturer to promote competition, and to revive the declining spirit of adventure by making intensive long-term investments by choosing areas that Korea can uniquely dominate.[2] The government has identified 30+ regulatory issues, spanning vehicle standards, operational rules, data usage and infrastructure, that require resolution for smooth AV deployment. Autonomous driving beyond Level 3 still faces technological barriers, as nighttime driving conditions, with less light than daytime, pose greater challenges for autonomous vehicles. Hyundai Motor, a Korean automaker, had originally planned to add a Level 3 autonomous system to its Genesis G90 by the end of 2024. However, the company has postponed that launch indefinitely. Vehicles with Level 3 technology, also known as "conditionally automated driving,” can monitor their environments and manage most aspects of driving without human assistance, though they will occasionally request human intervention. Korean automakers are finding difficulties in transitioning from a hardware-centric automaker to a software-defined vehicle (SDV) and AI-driven mobility companies. The companies will continue evaluations until they can be 100 percent confident in driver safety. From 2020 to September of 2025, there were a total of 112 traffic accidents involving autonomous vehicles operating on temporary permits. The highest number occurred in 2025, with 47 incidents as of September 2025, followed by 31 incidents in 2024, 27 incidents in 2023 and seven in 2022. These accidents may impact the public’s trust in current autonomous vehicles technology and limit progress until technological improvements occur to further reduce accidents. Obtaining permits to operate and test autonomous vehicles has been challenging. However, MOLIT is implementing a streamlined permit system for temporary operation of autonomous vehicles and expanding the scope for recognizing identical autonomous vehicles. Notably, it is improving regulations to extend the current five-year temporary permit period by two additional two-year extensions, allowing for a maximum duration of up to nine years.
Pursuant to Article 28-2 of the Personal Information Protection Act and the Personal Information Processing Guidelines for Autonomous Vehicles, data subjects are entitled to exercise their right to data portability, allowing them to request the direct transfer of diverse in-vehicle driving data to third-party service providers of their choice.[3] Guidelines for cybersecurity The Korean government issued guidelines based on international standards for automobile cyber security in December 2020 as a first step in addressing cybersecurity risks. The government stated it planned to enact laws/legal standards regarding automobile cybersecurity, with the goal of implementing such laws. On January 25, 2024, MOLIT amended the Motor Vehicle Management Act to establish automakers' obligations regarding cybersecurity management systems and software update safety measures for vehicles, as well as the authority of the Ministry as a relevant regulatory body. This amendment lays the institutional foundation for incorporating the international standards on automotive cybersecurity established in 2020. Amended Motor Vehicle Management Act (effective August 2025) Starting on August 14, 2025, according to MOLIT Ordinance No. 1519, South Korea’s revised Automobile Management Act requires manufacturers to establish cybersecurity management systems and obtain certification from MOLIT for self-certification. The updated regulations apply to most vehicles, excluding motorcycles, certain trailers and emergency vehicles. The cybersecurity management system signifies the organization, means and procedures for monitoring cyber threats at all times from vehicle production to operation and responding quickly to threats when they occur. MOLIT may investigate the compliance of autonomous vehicle manufacturers with software updates through a performance testing agency and order corrective measures for inappropriate updates. Global information security standard Hyundai Motor and Kia are strengthening in-car cybersecurity features. They have earned certifications of their cyber security and management system (CSMS) as meeting the United Nations Economic Commission for Europe (UNECE) regulation R-155 for new car releases as of July 2022. They changed the work process for establishment and operation of the CSMS during the entire life cycle of cars. LG Electronics received TISAX (Trusted Information Security Assessment Exchange) certification in all major areas of its electronic devices business to strengthen competitiveness in the automotive parts business. TISAX is a global information security certification created by a German automobile industry association to standardize the security evaluation criteria of different automobile manufacturers. It evaluates security in four aspects: information security system, partner security system, data protection systems and prototype protection systems.
Hyundai Mobis has announced plans to develop an embedded in-vehicle communication module capable of 5G wireless connectivity to support the industry's shift toward Software-Defined Vehicles (SDVs). The introduction of 5G technology is expected to enable practical remote control capabilities, allowing human operators to intervene in complex road situations, while also ensuring the stable implementation of high-precision map streaming services that exchange sensor data and road conditions in real time. Consequently, Hyundai Mobis aims to complete the development of this 5G-based telematics solution by the first half of 2026.[4] 5G+ Strategy In April 2019, the Korean government announced a “5G+ Strategy” to realize innovative growth by 2026 based on 5G technology, through cooperation among related departments and agencies. The Ministry of Science and ICT (MSIT) will support 5G V2X communications testbeds and verification of 5G autonomous driving performance on real roads. Current development/progress South Korea’s national ICT (Information, Communications and Technology) standards ecosystem continued to advance C-V2X standardization in 2025, with a focus on security architecture and testing frameworks for ultra-reliable, low-latency connected-vehicle services. The Telecommunications Technology Association (TTA) announced 2025 C-V2X security deployment and testing features to help translate 5G+ policy goals into procurable and verifiable technical requirements. LG Innotek announced a third-generation 5G Communication Module (Module) as an upgrade from earlier narrowband satellite modules to broadband 5G satellite.[5] The upgrade enables vehicles to seamlessly send and receive large volumes of data at high speeds, making advanced communication services a reality. Furthermore, the Module enhances autonomous driving performance, resulting in improved driving accuracy and safety. It adopts the latest 5G standard, Release 17 from 3rd Generation partnership Project (SGPP)[6]. On November 12, 2024, LG Electronics introduced its new cockpit concept for future mobility, called the “Digital Cockpit Gamma.” It consists of three main modules: Vision Display, Intelligent HMI and Connectivity & Content. It features a modular system that allows users to select their desired functions.[7] The Vision Display uses a transparent OLED instrument panel to provide essential driving information. The Intelligent HMI integrates artificial intelligence (AI) to improve driver interaction, including real-time monitoring that can offer smart services, such as suggesting nearby cafes for coffee if it detects drowsy driving. Connectivity & Content is characterized by 5G communication for fast data exchange and in-car entertainment via a webOS content platform in the car, enabling high-definition content and live broadcasts. The digital cockpit is part of the Mobility Labworks series, dedicated to researching the future of mobility with a focus on integrating innovative technologies in transportation. LG collaborates with car manufacturers to offer systems adapted to industry and user requirements. The company continues to strengthen its position in the automotive technology market and to support its broader commitment to the future of mobility.
[8] 외국어번역 > 영어번역 > 도로교통법 | 국가법령정보센터(The Traffic Laws: National Legal Information Center) [9] 외국어번역 > 영어번역 > 도로교통법 | 국가법령정보센터(The Traffic Laws: National Legal Information Center) [10] See Article 11 and 12 of the Act on The Support For The Innovation and Revitalization of Mobility. [11] See Article 7(2) of the Autonomous Vehicle Act. [12] See Article 9(2) of the Autonomous Vehicle Act. [13] These amendments are in step with the revised Autonomous Vehicles Act to establish the standards, targets, and methods for Performance Certification and Conformity Approval, required technical, managerial, and physical measures for approved operators, post-management obligations for those who obtain certification, and procedures for designating a dedicated agency. [14] See Article 56(2) of the Road Traffic Act. [15] Further details available at https://www.koroad.or.kr/main/board/6/301938/board_view.do?&listType=list&bdOpenYn=Y&bdNoticeYn=N CHECK LINK [16] See Article 145(2) of the Road Traffic Act.
Development requirements for self-driving cars and autonomous driving systems have been revised in accordance with the "Act on Promotion and Support of Commercialization of Self-driving Cars," which took effect on February 1, 2026. The changes are intended to systematize the "g Performance Certification" and "Conformity Approval" systems that allow vehicles without safety standards to operate in preparation for the advancement of autonomous driving technology (Level 4 or higher). And Article 108-2 (education related to autonomous vehicles) under the Road Traffic Act: For the safe operation of autonomous vehicles, as prescribed by Presidential Decree, those who have obtained temporary driving permission for autonomous vehicles must complete safety training before operation.[8] Act on The Support for the Innovation and Revitalization of Mobility (effective October 19, 2023) The Act on the Support for the Innovation and Revitalization of Mobility (Act) was enacted to support diverse and creative mobility services led by the private sector, as the integration of innovative technologies. The regulatory sandbox system included in the Act consists of two main components: a Regulatory Fast-Track System and an Experimental[9] Exception System.[10] Revised Autonomous Vehicle Act (effective July 10, 2024) A system for the discretionary designation of pilot operation zones that cover two or more cities and provinces was established.[11] The authority to grant permits for passenger transport businesses within the pilot operation zones has been transferred from MOLIT to the heads of local governments in respect of the pilot operation zones.[12] In 2025, MOLIT amended the Enforcement Decree of the Act on the Promotion of and Support for Commercialization of Autonomous Vehicles (Autonomous Vehicles Act) (Presidential Decree No. 35489) to accelerate the commercialization of self-driving cars and partially amended the Enforcement Rule (MOLIT Ordinance No. 1496) to build out the detailed legal framework for AV safety/approval.[13] Road Traffic Act Amendment (effective September 20, 2024) The amendment to the Road Traffic Act contains a requirement that drivers of partially autonomous vehicles respond to the autonomous driving system’s direct driving request without delay and directly operate the steering wheel or brakes.[14] From March 20, 2025, drivers operating under a temporary autonomous vehicle driving permit must complete mandatory safety and traffic education, with an administrative fine for noncompliance.[15] The commissioner of the national police agency may establish and operate a traffic information center to collect, analyze and provide traffic information.[16]
The Korean government has designated the Autonomous Vehicle Control Zone (AV POZ) and the mobility permit has requested fewer than 10 vehicles. This has limited the quantitative growth of autonomous driving technology. To overcome this, the Korean government has announced that it will increase the number of vehicles for pilot operation to 200. And SWM, which specializes in autonomous driving technology, said it aims to establish a cooperative model for future mobility with the Seoul Taxi Association. This agreement implies the possibility of establishing a Korean-style robotaxi cooperative model that can be integrated and expanded within the public transportation system. Finally, the government announced that Gwangju Metropolitan City has been selected as the test site for the first autonomous vehicle in Korea. According to the Ministry of Land, Infrastructure and Transport, the entire road network in Gwangju will be opened as a test site for autonomous vehicles.[17] Hyundai Motor Group and Kia will launch Its SDV pace car in 2026 In 2026, Kia plans to unveil its Software Defined Vehicle (SDV) Pace Car, featuring full-stack SDV technology and AI-integrated autonomous driving technology. After this debut, Kia aims to introduce more advanced and reliable Level 2+ functions and expand AI-based features. Kia is also preparing for mass production and a broader software ecosystem beginning in 2027. Hyundai Motor Group set out a 2026-focused strategy. HMG highlighted its work on AI-powered autonomous driving, using an end-to-end deep learning model (Atria AI), developed with 42dot and Motional. HMG framed 2026 as a year where its enabling SDV/AI architecture is expected to be demonstrated more concretely. Goals set out by South Korea to be achieved by 2026 The largest change South Korea is publicly signaling is a move from route-by-route pilots to city-scale deployment of autonomous vehicles in Gwangju City. MOLIT intends to prepare a large city-scale deployment framework through preliminary research and to finalize the master plan in 2026. Construction will begin in 2028, aiming to be completed by the end of 2030. The pilot city will be created as space where advanced mobility technologies such as autonomous vehicles, robots, drones and UAM are applied to everyday life. It further describes broader support measures such as regulatory rationalization and an AI learning center. To support the AI Transformation and Ultra-Innovation Economy flagship initiatives, the government intends to back the initiatives with substantial fiscal investment and sweeping regulatory reforms. As an initial priority, it will promote autonomous vehicles as a key “physical AI” industry, with the goal of commercializing AVs fully by 2027. Starting in 2026, the government will designate autonomous-driving demonstration cities, where autonomous vehicles can operate citywide to accelerate the collection of large-scale driving data. The government aims to operate more than 100 self-driving vehicles citywide in Seoul and to commercialize Level 4 autonomy by 2027. The plan includes expanding beyond the current 47 pilot zones and supporting dedicated Graphics Processing Units (GPUs) for autonomous vehicles. Furthermore, the plan would allow the use of original video data and reduce the need for manual blurring that previously raised privacy concerns. Once anonymized or pseudonymized, video data gathered from privately owned vehicles will be permitted for research and development. It also calls for expanding Level 3 autonomous buses in rural areas with limited transportation options.
Last year, Samsung C&T, in partnership with the autonomous robot company Neubility, conducted a pilot project at the Raemian Leaders One apartment complex in Seocho-gu, Seoul, to see if robots could deliver food to individual units in residential complexes. In 2026, the service area has expanded beyond the area near Raemian Leaders One, and through integration with the food delivery platform Yogiyo, about 130 restaurants and cafes within a 1.2 km radius have been connected.[18] Government’s Supportive Stance South Korea’s AV efforts extend well beyond passenger cars, encompassing freight transport, delivery services, farming and more, as the country seeks to lead in diverse autonomous industries. In June 2021, the Korea Transportation Safety Authority established the AV Transportation and Logistics Master Plan 2025 to commercialize autonomous driving-based transportation and logistics systems through the commercialization and proliferation of autonomous vehicles. Hyundai Motor and Kia’s DAL-e Delivery robot arrives, ready to revolutionize indoor autonomous delivery services Hyundai Motor and Kia's new autonomous delivery robot, DAL-e Delivery, provides swift and reliable delivery services in complex indoor environments.[19] DAL-e Delivery is equipped with four Plug & Drive (PnD) modules, a mobility solution that combines a motor with steering, suspension, braking systems and environmental recognition sensors. The key feature of DAL-e Delivery is its autonomous driving capability to navigate and transfer to and from all building floors, seamlessly interfacing with the elevator and door control system. It also ensures quick delivery service by creating optimal routes in real-time. CJ Logistics tests autonomous parcel trucking CJ Logistics, a leading supply chain and technology innovator, plans to expand logistics automation by introducing autonomous driving technology for its delivery trucks.[20] The highway freight autonomy is moving from a company-by-company pilot concept toward system-level scale-up. The nationwide expressway pilot-zone expansion in March 2025 materially enlarged the operational envelop from hub-to-hub tracking services. Furthermore, in November 2025, a program for unmanned autonomous driving commercialization for large-truck cargo transport to be led by MARS AUTO from September 2025 to 2027 was announced.[21] CJ Logistics’ autonomous driving trucks operate manually in urban areas and switch to autonomous driving on the highway. CJ Logistics anticipates that autonomous driving will reduce truck drivers’ workloads and enhance safety. South Korea is investing in “smart farming” innovations that include autonomous machinery. The Ministry of Agriculture, Food and Rural Affairs (MAFRA) announced its First Five-Year Master Plan (2025-2029) aimed at promotion of the growth of the smart-farming industry and advancing autonomous tractors and robotic farm equipment to address rural labor shortages, climate change and the need to modernize farming through CIT and automation.[22] Testing standards for autonomous agricultural machines are being established, and subsidies are being provided to encourage adoption of self-driving tractors and drone technologies on farms.
[23] 외국어번역 > 영어번역 > 자율주행자동차 상용화 촉진 및 지원에 관한 법률 | 국가법령정보센터(Act on Promotion and Support of Commercialization of Self-Driving Vehicles: National Legal Information Center), Korea sets AI safety rules, first in world - The Korea Herald [24] Article 25(2) of PIPA regulates mobile video processing for business purposes in public areas [25] Article 25(2) of PIPA regulates mobile video processing for business purposes in public areas [26] Further details available https://www.chosun.com/english/industry-en/2025/11/17/EXLEHHMUHZAETN43YOI5W7KMPY/ [27] Further details available INFORMATION AND LINK ARE MISSING
AI Framework Act (Effective January 2026): This Act designates autonomous vehicles as “High-impact AI,” mandating rigorous safety standards while providing transparency and reliability-based guidelines to enable technological development free from legal uncertainty.[23] The Personal Information Protection Commission (PIPC) releases guidelines The PIPC’s guide establishes specific standards for using personal information collected by mobile devices operating in public spaces.[24] It also stipulates obligations under Article 25(2) of the Personal Information Protection Act (PIPA).[25] For video captured by rapidly moving devices, such as autonomous vehicles, operators are recommended to delete unnecessary footage after use and anonymize or pseudonymize the data to reduce access or deletion requests. Using footage recorded during test operations for autonomous vehicle technology research, without pseudonymization or anonymization, is likely not foreseeable to data subjects. On August 6, 2025, the PIPC published a Generative-AI (GenAI) personal data processing guide, explicitly aimed at reducing uncertainty under the PIPA for companies developing or using GenAI. Driverless vehicle testing and deployment STRADVISION, a trailblazer in deep learning-based vision perception technology for the automotive industry, in May 2023 established an “Autonomous Driving Workshop” in Dongtan, Gyeonggi-do, South Korea. This workshop was focused on enhancing vehicle object recognition technology and boasted facilities for testing and advancing camera-based autonomous driving, as well as other sophisticated technologies such as LiDAR and radar. Current AI laws and deployment – AI Basic Act AI-related legislation was proposed in the 21st National Assembly beginning in 2020 and passed the National Assembly in December 2024, taking effect in January 2026. The law sets overarching principles for safe and trustworthy AI development, such as requirements for transparency and protection of users’ rights. Draft enforcement guidelines released in late 2025 highlight voluntary AI impact assessments and government support for AI R&D rather than heavy-handed restrictions.[26] For the AV industry, this means core vehicle AI algorithms will need to align with the new law’s criteria. The framework may accelerate development, ensuring autonomous vehicles become more AI-driven under a clear national policy. The government expands financial support for AI cips The Ministry of Trade, Industry and Energy (MOTIE) is expanding financial support for the development of AI semiconductors to strengthen the competitiveness of the industry. MOTIE plans to develop a generic, open next-generation vehicle AI accelerator semiconductor for SDVs, capable of 1,000 trillion operations per second (TOPS).[27] Also, it aims to pursue a high-speed AV network system and core chips capable of 10 Gbps to support Level 4 autonomous driving.
[2] Conditions for Korean self-driving cars’ success
[17] Autonomous Vehicle, Weekly News #04 / 2026 | by Youngseop Song | Jan, 2026 | Medium, Gwangju becomes Korea's 1st test bed for self-driving cars - The Korea Times
[18] Robots deliver meals door-to-door in Seoul apartments | QSR Media Asia [19] Further details available at https://www.hyundai.com/worldwide/en/newsroom/detail/hyundai-motor-and-kia%25E2%2580%2599s-dal-e-delivery-robot-arrives%252C-ready-to-revolutionize-indoor-autonomous-delivery-services-0000000716. [20] Further details available at https://newsroom.cj.net/cj-logistics-tests-autonomous-parcel-trucking/ [21] Further details available at https://www.chosun.com/english/industry-en/2025/11/17/EXLEHHMUHZAETN43YOI5W7KMPY/ [22] https://www.mafra.go.kr/bbs/english/25/573412/artclView.do
[3] 외국어번역 > 영어번역 > 개인정보 보호법 | 국가법령정보센터 (Personal Information Protection Act:National Legal Information Center)
[4] Hyundai Mobis to develop 5G telematics for vehicles - 매일경제 영문뉴스 펄스(Pulse) [5] Further details available at https://www.lgcorp.com/media/release/29050 [6] Release 17 is the first international standard to extend 5G technology from traditional terrestrial networks to non-terrestrial networks, such as satellites. (Further details available at https://www.3gpp.org/specifications-technologies/releases/release-17) [7] Further details available at https://www.mk.co.kr/en/business/11164745.
The UK government continues to position autonomous vehicles as a major driver of economic growth and innovation over the coming decade, predicting that self-driving technology will support the creation of 38,000 new high-skilled jobs, and contribute more than £42 billion to the UK economy in 2035. Against this backdrop, 2025 was a pivotal year for regulatory and policy development, and 2026 has continued in a similar manner. A key announcement has been the government's decision to accelerate the introduction of Automated Passenger Services (APS). Subject to the outcome of ongoing and upcoming consultations, commercial pilot services carrying paying passengers without safety drivers are anticipated to begin operating over the course of this year. These pilots are intended to cover taxi- and bus-like services operating on public roads. To support these ambitions, the UK has taken significant legislative steps. Following extensive industry engagement and consultation, which included the publication of the Law Commission’s Joint Report on Autonomous Vehicles (Joint Report) and Connected & Automated Mobility 2025: Realising the benefits of self-driving vehicles in the UK by the UK government in August 2022 (Mobility 2025), the government followed this with the Automated Vehicles Act 2024 (AV Act). The AV Act establishes a comprehensive legal framework to enable the safe deployment of autonomous self-driving vehicles on public roads in the UK. It implements many of the recommendations made by the Law Commission in its Joint Report. However, much of the detailed regulatory framework required to implement the AV Act in practice will be introduced through secondary legislation and regulatory guidance, with the UK government currently aiming to implement the wider regulatory regime in the second half of 2027. Pending the introduction of the wider regulatory framework under the Automated Vehicles Act 2024, the Department for Transport and the Centre for Connected and Autonomous Vehicles (CCAV) published guidance in March 2026 establishing a new self-driving vehicle pilot scheme. The scheme is intended to support advanced trials and early commercial deployment on public roads while helping regulators and industry gather operational evidence ahead of the remaining provisions of the AV Act coming into force, which is currently expected in late 2027. This pilot scheme and related guidance are considered in further detail in Question #2 below. Call for evidence 2026 In late 2025, the UK government launched a call for evidence to inform secondary legislation and regulatory guidance under the AV Act. The consultation, which closed on March 5, 2026, contained 125 questions and sought views on, among other things: Type approval, authorization, user-in-charge and transition demands, no-user-in-charge operator licensing, insurance, data collection, cyber security, in-house regulation, monetary penalties, and incident investigation. The consultation aimed to ensure that regulatory framework both embeds robust safety requirements and remains flexible enough to adapt as technology evolves. A government response and summary of consultation feedback will be published, followed by a further consultation in the second half of 2026. Current timelines indicate that the core regime of regulations supporting large-scale autonomous vehicle deployment will be in place in the second half of 2027. First driverless vehicles to be launched in the UK in 2026 In parallel with regulatory development, industry momentum is accelerating. Several major robotaxi operators have announced plans to commence operations in London from April 2026, subject to regulatory approval. Waymo launched trials on public roads in London in April 2026, with the company reportedly targeting commercial passenger operations from late 2026, subject to regulatory approval. Lyft and Uber have also announced their plans to expand driverless operations into London this year through international partnerships. London-based Wayve Technologies has also developed a cutting-edge Driver AI system and has partnered with Uber to launch fully autonomous vehicles in London.
While the UK is generally regarded as a supportive jurisdiction for autonomous vehicle development, there are several regulatory and practical constraints that developers must navigate. Driverless Testing and Deployment Testing without a safety driver The Department for Transport and the Centre for Connected and Autonomous Vehicles (CCAV) published guidance on the UK’s new “self-driving vehicle pilot scheme” on March 31, 2026. The guidance applies across England, Scotland and Wales and sets out the framework for organizations seeking to operate autonomous vehicles on public roads without a human safety driver present in the vehicle. The pilot scheme is intended to support real-world testing of automated vehicle technology ahead of the full implementation of the Automated Vehicles Act 2024 (AV Act). It is designed to enable government, regulators and industry to gather operational evidence relating to safety, passenger services, incident response and regulatory oversight in live road conditions. This pilot scheme will also help inform the development of the permanent regulatory framework for automated vehicles in the UK. At a high level, the guidance imposes several categories of requirements, including: Vehicle approval requirements, including any required listing or authorization requirements under the UK automated vehicle framework and, where necessary, obtaining a "Vehicle Special Order" where the vehicle does not comply with existing construction and use requirements. Operational and safety requirements, including operational design domains, safety management systems, incident response procedures, cyber security and data management, passenger safety arrangements and engagement with emergency services. Reporting and governance obligations, including incident reporting, monitoring obligations and ongoing compliance with conditions imposed by regulators. Passenger service requirements, where the trial involves carrying passengers. An Automated Passenger Services (APS) permit may be required for taxi, or private hire vehicle-like services in England, or bus-like services operating in the UK. APS permits are granted by the Driver and Vehicle Standards Agency on behalf of the Secretary of State for Transport, and applicants may also need to obtain any relevant local authority consents or approvals before passenger services can commence. An organization seeking to operate a trial under the pilot scheme would therefore need to prepare a detailed application demonstrating that the proposed deployment can operate safely and lawfully. In practice, applicants are likely to need to secure the relevant vehicle approvals, prepare operational and incident management documentation, engage with regulators and local authorities and, where passengers are carried, obtain an APS permit before deployment can commence. Testing with a safety driver The above noted pilot scheme does not apply to autonomous technology trials with a safety driver. Such testing activities are still expected to follow the Code of Practice: Automated Vehicle Trialing, published jointly by CCAV, the Department for Transport and the former Department for Business, Energy & Industrial Strategy. The code sets out best practices for safety, risk management, cybersecurity, data protection and public engagement. While the Code of Practice is not legally binding, failure to follow it may be taken into account by courts and regulators when determining liability following an incident. In practice, this creates a strong incentive for compliance. To test an AV on a UK road under the Code of Practice, a driver or operator who can resume control of the vehicle must be present, the vehicle must be roadworthy and appropriate insurance must be in place. The code notes that anyone planning to conduct such a trial should consult the road and enforcement authorities, develop engagement plans and have data recorders fitted. Those planning to conduct advanced trials should contact the Centre for Connected and Autonomous Vehicles (CCAV) in advance. Vehicle standards Prior to entering the UK market, any autonomous vehicle must comply with the type approval regime, which assesses whether the vehicle meets all applicable technical, safety, environmental and other performance standards. In addition, vehicles intended to operate without a driver must undergo a separate authorization process under the AV Act to demonstrate they can drive safely and legally on UK roads. Guidance on compliance with these regimes has yet to be published, and secondary legislation is required to determine what vehicles are suitable to be used as "self-driving vehicles."
Autonomous vehicles rely inherently on the large-scale collection and processing of data, including location data, sensor data, telematics and user information. Much of this information constitutes personal data under the UK GDPR, the Data Protection Act 2018, the Data (Use and Access) Act 2025 and the Privacy and Electronic Communications Regulations. Although these laws pre-date widespread deployment of autonomous vehicles, they are technology-neutral and fully applicable. The Information Commissioner’s Office (ICO) has published guidance in response to the Law Commission’s Joint Report, clarifying how data protection law applies to automated and self-driving vehicles. Key themes from the ICO’s guidance include: Data protection by design and by default: privacy considerations should be embedded from the earliest stages of system design. Data minimization: systems should only collect personal data that is strictly necessary, with particular caution around continuous location tracking. Transparency: controllers must provide clear and accessible privacy information, including in scenarios where data is collected from individuals with no direct relationship to the service provider (e.g. pedestrians captured by sensors). Local processing and anonymization: where possible, personal data should be processed within the vehicle itself and anonymized or pseudonymized before transfer. Security: controllers must implement appropriate technical and organizational measures, including encryption, access controls, system segregation and prompt patching of vulnerabilities. The Department of Transport also provided cyber security guidance for autonomous vehicles to embed security practices at every level of design, construction, supply, support and operation in 2018. The UK government also recommends several ISO standards be met, from security techniques and software testing to privacy and handling personally identifiable information. Governance: data protection impact assessments (DPIAs) are expected where processing is likely to result in high risk to individuals. For data protection purposes, the entity operating the autonomous driving system and providing the service will typically be the data controller. This means the operator carries primary legal responsibility for ensuring all personal data processed by the vehicle or service complies with data protection law; it cannot shift that accountability to suppliers or technology partners, even where the processing is technically carried out by others.
Extension of 5G coverage Reliable, low-latency connectivity is a critical enabler for autonomous vehicles, particularly for remote supervision and intervention. In 2025, Ofcom reported that outdoor 5G coverage from at least one operator now reaches approximately 97 percent of the UK, representing a significant increase from 2024. Coverage varies between networks, but continues to expand yearly. In Ofcom's Plan of Work 2025/2026 Consultation, Ofcom identified 6G and advanced wireless technologies as an area of strategic focus, noting their potential to support robotics, automation and future mobility technologies through improved responsiveness, capacity and security. 5G-enabled Connected and Automated Logistics (CAL) The 5G CAL project, led by the North East Automotive Alliance (NEAA), has demonstrated how high-speed, low-latency networks can support autonomous logistics, including autonomous HGVs operating on private roads. The project has been particularly significant in enabling remote operation without a safety driver, including for vehicles weighing up to 40 tons. The program has helped overcome one of the key barriers to autonomous freight operations and has informed subsequent logistics automation initiatives.
The most significant legal development is the Automated Vehicles Act 2024, which substantially expands the UK’s regulatory framework beyond the Automated and Electric Vehicles Act 2018 (AEVA). Key features include: Authorization regime for self-driving vehicles operating on public roads. Introduction of Authorized Self-Driving Entities (ASDEs), responsible for ongoing compliance and safety assurance. A statutory self-driving test, requiring vehicles to operate autonomously, safely and legally. Clear allocation of responsibility between user-in-charge and no-user-in-charge models. Enhanced insurance and liability framework, with insurers compensating victims first and seeking recovery within the liability chain. A new operator licensing regime for no-user-in-charge services. Extensive information-gathering, monitoring and investigation powers. Civil sanctions, new criminal offenses and enforcement mechanisms. Marketing restrictions to prevent consumer confusion. A dedicated regime for Automated Passenger Services, disapplying certain taxi and bus laws where appropriate. While much detail will be set out in secondary legislation, the AV Act provides long-awaited regulatory clarity and is widely seen as a catalyst for investment. Secondary legislation is currently undergoing consultation, with the expectation of being released in the later half of 2027, which should fill any gaps in the current legislative framework.
Transport authorities are working closely with operators to ensure safety, local engagement and regulatory compliance. 2026 is expected to mark the UK’s transition from testing to early commercial deployment. The government’s accelerated APS program is intended to enable driverless taxi and bus-like pilots in England from Spring 2026. Waymo has announced plans to launch a pilot robotaxi service in London in April 2026, with paying passengers expected by September 2026, subject to regulatory approvals. Waymo already operates engineering hubs in London and Oxford and brings extensive operational experience from the United States. Competition is intensifying. Uber and Lyft have announced plans to deploy robotaxis in London through partnerships with Baidu’s Apollo Go, one of the world’s most mature robotaxi platforms. Subject to approval, this would represent one of the first large-scale deployments of Chinese autonomous vehicle technology in a European capital.
Public Transport StreetCAV Plus – Milton Keynes StreetCAV Plus is a city council-led project deploying low-speed autonomous shuttles in Milton Keynes in cooperation with local company Smart City Consultancy. The vehicles, built by New Zealand company Ohmio, are fully accessible and designed to carry up to eight passengers. Mapping work began in early 2025, and trials will initially operate at speeds below 15 mph. The project has already created skilled local jobs and is intended to pave the way for future driverless deployment. With new funding in place, testing and trialing can expand to wider parts of the city, including the ongoing trial between the train station and Hotel La Tour. Tourism Shuttles – Orkney In the Scottish island of Orkney, a driverless vehicle project led by Urban Foresight and autonomous shuttle firm Aurrigo and transport body Hitrans has the plan to provide a shuttle service between Kirkwall Airport and Harbour. However, their current focus is on connecting some of the historic monuments across the island. Concerns such as the dramatic weather faced on the island were raised and the company flagged that there are "a few more steps" to meet prior to running vehicles on the full route. Their current study ended in March 2026, and there are plans for a full trial in late 2026 or 2027. Commercial Trucking DHL Supply Chain – Heathrow Airport DHL Supply Chain has successfully completed a live trial of autonomous vehicles at Heathrow Airport, covering more than 1,300 km in operational airside traffic across 14 days. The long-term ambition is to automate baggage transfer operations using electric autonomous vehicles, improving efficiency and reducing emissions. Currently, DHL's baggage transfer services handle the transfer of more than 60,000 passenger bags between terminals each month. The initial project has been conducted using a passenger vehicle to demonstrate the capabilities of the technology, but this will ultimately be improved to include electric vans and tow tractors suitable for baggage transfer. Oxa – Autonomous logistics and ports As part of the £150 million CAM Pathfinder program, Oxa and NEAA are delivering P-CAL (Port-Connected and Automated Logistics), deploying autonomous terminal tractors and secure communications networks at UK ports. The project aims to address the complexities of commercializing CAM vehicles to support and grow the UK's CAM supply chain. By creating a UK first in waterside port automation, this initiative will deploy a fully autonomous terminal tractor and secure mesh communication network to move containers between the dockside and the container compound. The project builds on earlier 5G CAL and V-CAL initiatives and aims to support Net Zero objectives while improving safety and efficiency. Self-driving trucks – Suffolk In September 2025, the Port of Felixstowe in Suffolk brought 34 autonomous trucks from Shanghai Westwell Technology Company, doubling the number of autonomous vehicles already in operation there. The use of battery-powered trucks is now possible following the introduction of a private 5G network at the port, and the installation of an automated battery swap station allows power units to be changed to fully charged ones in minutes without manual intervention. The vehicles use 360-degree vision and monocular and stereo cameras with radar alongside conventional trucks requiring drivers. The port has declared that no jobs will be at risk as a result of the increase in automated vehicles. The port is proud to be at the front of autonomous technology and aims to be more efficient and resilient in its operations and achieve net-zero by 2035. Autonomous Food Delivery Just Eat – Milton Keynes Just Eat has started a small-scale UK trial of autonomous delivery robots, deploying machines from Delivers.AI in Bristol and RIVR robots in Milton Keynes to bring food from local restaurants straight to customer doors. The move is part of a broader exploration of ground robotics for last-mile delivery, following earlier pilots in Switzerland where nearly 1,000 autonomous deliveries were completed. In Milton Keynes, the RIVR machines are designed to handle urban obstacles and go right to the doorstep. This trial sits alongside other initiatives in ground robots and separate drone trials in Ireland, with the UK pushing autonomous delivery beyond roads into everyday services. The trial sees the broader trend in the UK toward testing and scaling autonomous systems for real-world use. This technology is not new to Milton Keynes, with Starship Technologies launching its pilot commercial service in the city in 2018. Uber Eats and Co-Op – Leeds Since 2022, the Co-Op and Starship Technologies have partnered to deliver groceries to the residents of Leeds, using six-wheeled autonomous delivery robots. Since January 2026, Starship has also joined with Uber Eats, expanding this delivery service. The robots are able to complete deliveries in less than 30 minutes for a distance of up to two miles. The robots are aimed at reducing congestion from short trips and improving air quality. In relation to safety, concerns have been raised for the inhabitants of Leeds; however, Starship has confirmed that the machines have completed more than nine million deliveries across seven countries and Leeds has welcomed the future of autonomous deliveries. Autonomous drones – area to watch A leading global technology and e-commerce company has successfully started drone flights from their Darlington fulfilment center, with the city becoming the first UK location to launch Prime Air later in 2026. The services are to use Amazon's latest MK30 drone, which can operate both safely and autonomously using state-of-the-art detect-and-avoid technology to ensure the safety of people and property. Prime Air is designed to safely deliver packages weighing up to five pounds to customers in under two hours. Over the course of 2026, once the platform is launched, eligible Prime Air customers based in the area can opt-in to drone delivery and choose from thousands of products to be delivered to them faster than ever before.
AI laws The UK government has considered a non-statutory approach to AI, with the knowledge that this offers “critical adaptability” to a consistently changing landscape. Many of the UK AI laws are set out across the UK GDPR and the Data Protection Act 2018, the Online Safety Act 2023 and the government's white paper (2024). At present, AI-specific legislation has not materially constrained autonomous vehicle deployment in the UK. The UK continues to favor a principles-based, sector-led approach to AI regulation, with existing laws and regulatory frameworks (including those relating to safety, data protection and product liability) applying to AI systems within the remit of existing regulators. Unlike the EU, the UK has not introduced a dedicated AI Act adopting a comprehensive risk-tier framework or blanket prohibited categories, although the government has indicated that targeted legislative intervention may be introduced in specific high-risk areas as the technology and regulatory landscape develops. AI technology However, AI is increasingly central to autonomous vehicle development. Wayve, a UK company focused on developing cutting-edge AI technology to power autonomous vehicles, recently hailed the UK's "AI ecosystem" as having been a key factor in the company securing the largest ever investment in a UK AI company during its most recent fundraising round. It focuses on developing its own AI-driven autonomy. Wayve is key in this emerging landscape by partnering with Uber on UK robotaxi trials, aiming to leverage machine-learning-based driving technology on city streets. Due to developing and uncertain legislative frameworks in the UK regarding AI, companies operating in this space need to be mindful of future legislative developments, as AI will play a key role in autonomous technology. Transport AI action plan: transforming ambitions In June 2025, the UK government launched its Transport AI action plan: transforming ambitions. Following on the heels of the 2025 AI opportunities action plan, this plan sets out the Department of Transport's approach to working with AI, including the implementation of AI in the department to support businesses deploying AI across the transport system to improve transport user experiences and boost growth. The plan envisions responsible, safe, secure and transparent AI across the transport system, including those for autonomous vehicles with an explicit aim of improving safety, efficiency and user experience. It includes actions that will enable advanced trials and early commercial pilots of AI-enabled and AV technologies ahead of the regulator regimes fully coming into force.
