1. Drive a university-wide approach
2. Invest in organisational enablers
3. Trial and learn in partnership
1. Drive a university-wide approach
1A
Incorporate AI into university strategy.
Universities should map the opportunities and threats AI technologies present to their existing strategies. They can use this to articulate the objectives for AI to reinforce strengths (such as data assets) and to address weaknesses. Understanding current organisational AI maturity is necessary to keep these objectives realistic.
1B
Develop or adopt overarching AI principles to guide implementation.
Universities should establish ethical guidance for the use of AI in student services. They should ensure transparency, fairness, and accountability in AI-driven decision-making processes, and regularly review and update these principles to align with evolving best practices (for example the Russell Group’s recently published principles on the use of generative AI tools in education).
1C
Prioritise AI investment based on known user needs.
Universities should resist the temptation to take a technology or product-led approach to using AI applications. Adoption of AI tools should instead address evidenced pain points where interventions will add most value from a student experience perspective.
2. Invest in organisational enablers
3. Trial and learn in partnership
Empower leaders to experiment.
Within strategic and ethical guardrails, university leaders should be given licence to creatively explore potential use cases for AI to find out what works through hands-on trial and error. A relatively modest budget to deploy applications in limited or sandbox environments can help to develop proof-of-concept while demystifying AI for the staff involved.
1D
Invest in data infrastructure.
Robust student data collection, storage and management systems are the backbone of AI applications in education. AI thrives on access to large datasets and diverse data inputs. Universities must therefore invest in building scalable infrastructure to securely handle increasing volumes of data. This infrastructure needs the ability to integrate various data sources (including internal systems, data from external providers, and campus IoT devices) to power more accurate and comprehensive insights.
2A
Build AI literacy in staff, faculty, and students.
Provide training for university staff and faculty to help them understand AI's capabilities and limitations, including on the ethical and responsible use of AI. This empowers educators to integrate AI into their teaching methods and to advise students, who will learn how to use AI appropriately while still developing critical thinking and reasoning skills.
2B
Prioritise accessibility and inclusivity.
Ensure that AI tools are designed with accessibility and inclusivity in mind. Conduct regular accessibility audits to identify and rectify any barriers that may exclude certain students from benefiting fully from AI-driven services.
2C
Collaborate with industry experts and innovators.
Partnering with educational technology companies can provide universities with access to state-of-the-art AI tools and expertise. Collaborative ventures can also help tailor AI solutions to the specific needs of a university. Developing checklists of considerations for vendors to know about before entering into agreements with campuses can also be valuable. Universities can also look internally to AI expertise held in their own academic workforces.
3A
Cooperate with competitors for economies of scale.
AI models are often most effective when trained on large datasets, which smaller institutions can struggle to provide. Universities should consider the strategic merits of cooperating with other institutions to combine their data assets, where there is a clear benefit to be derived for students. Partnership arrangements like this could also be a way to spread the upfront cost of procuring AI systems.
3B
Establish and use feedback mechanisms.
Implement feedback mechanisms for students, staff and faculty to voice their concerns and suggestions regarding AI-driven services. Continuous improvement based on user feedback is essential for optimising the student experience. This can help to ensure that AI tools are solving genuine problems, rather than being just technology for technology’s sake.
3C
Bring faculty, staff and students along for the journey.
Universities should adopt change management strategies to support implementation of AI tools (as well as other technological solutions). Demonstrating tangible applications of AI tools – including what they mean for different groups – helps to bring the university community along for the journey and reduce fatigue that comes from excessive AI hype.
3D
1D
Empower leaders to experiment.
Within strategic and ethical guardrails, university leaders should be given licence to creatively explore potential use cases for AI to find out what works through hands-on trial and error. A relatively modest budget to deploy applications in limited or sandbox environments can help to develop proof-of-concept while demystifying AI for the staff involved.
2A
Invest in data infrastructure.
Robust student data collection, storage and management systems are the backbone of AI applications in education. AI thrives on access to large datasets and diverse data inputs. Universities must therefore invest in building scalable infrastructure to securely handle increasing volumes of data. This infrastructure needs the ability to integrate various data sources (including internal systems, data from external providers, and campus IoT devices) to power more accurate and comprehensive insights.
2B
Build AI literacy in staff, faculty, and students.
Provide training for university staff and faculty to help them understand AI's capabilities and limitations, including on the ethical and responsible use of AI. This empowers educators to integrate AI into their teaching methods and to advise students, who will learn how to use AI appropriately while still developing critical thinking and reasoning skills.
2C
Prioritise accessibility and inclusivity.
Ensure that AI tools are designed with accessibility and inclusivity in mind. Conduct regular accessibility audits to identify and rectify any barriers that may exclude certain students from benefiting fully from AI-driven services.
3A
Collaborate with industry experts and innovators.
Partnering with educational technology companies can provide universities with access to state-of-the-art AI tools and expertise. Collaborative ventures can also help tailor AI solutions to the specific needs of a university. Developing checklists of considerations for vendors to know about before entering into agreements with campuses can also be valuable. Universities can also look internally to AI expertise held in their own academic workforces.
3B
Cooperate with competitors for economies of scale.
AI models are often most effective when trained on large datasets, which smaller institutions can struggle to provide. Universities should consider the strategic merits of cooperating with other institutions to combine their data assets, where there is a clear benefit to be derived for students. Partnership arrangements like this could also be a way to spread the upfront cost of procuring AI systems.
3C
Establish and use feedback mechanisms.
Implement feedback mechanisms for students, staff and faculty to voice their concerns and suggestions regarding AI-driven services. Continuous improvement based on user feedback is essential for optimising the student experience. This can help to ensure that AI tools are solving genuine problems, rather than being just technology for technology’s sake.
3D
Bring faculty, staff and students along for the journey.
Universities should adopt change management strategies to support implementation of AI tools (as well as other technological solutions). Demonstrating tangible applications of AI tools – including what they mean for different groups – helps to bring the university community along for the journey and reduce fatigue that comes from excessive AI hype.