There are four distinct stages of RAI maturity
Starting to implement an RAI program, with a focus on data and privacy
Expanding across
the remaining RAI dimensions and initiating RAI policies
Improving data
and privacy, but lagging behind on human-related advances
Performing at a high level across all RAI dimensions
Relative change in an organization’s focus
Stage 1
Lagging
Stage 2
Developing
Stage 3
Advanced
Stage 4
Leading
Accountability
Transparency and “explainability”
Fairness and equity
Safety, security, and robustness
Data and privacy governance
Social and environmental impact mitigation
Human plus AI
Accountability
Transparency and “explainability”
Fairness and equity
Data and privacy governance
Social and environmental
impact mitigation
Safety, security, and robustness
Human plus AI
Stage 1 to stage 2
Accountability
Transparency and “explainability”
Fairness and equity
Data and privacy governance
Social and environmental
impact mitigation
Safety, security, and robustness
Human plus AI
Stage 2 to stage 3
Accountability
Transparency and “explainability”
Fairness and equity
Data and privacy governance
Social and environmental
impact mitigation
Safety, security, and robustness
Human plus AI
Stage 3 to stage 4
Accountability
Transparency and “explainability”
Fairness and equity
Safety, security, and robustness
Data and privacy
governance
Social and environmental impact mitigation
Human plus AI
Seven generally accepted dimensions of RAI
Organizations and individuals designing, developing, deploying, or operating an AI system are responsible for the outcomes as well as for ensuring that they are used appropriately and effectively
Organizations and individuals designing, developing, deploying, or operating an AI system must be transparent, explaining, as required, the purpose and goals of the system, how it was developed, and how outcomes are reached
AI systems are designed to be inclusive, to identify and mitigate sources of bias, and to promote
fair outcomes
AI systems are designed to be secure and resilient and to have safeguards that reduce the risk of unintended behaviors and outcomes
AI systems are designed, and policies are in place, to ensure compliance with data privacy laws and to mitigate privacy risks
AI systems are designed to promote positive, sustainable impact and to avoid creating or perpetuating adverse effects on society and the environment
AI systems are designed to empower people (those who are developing, deploying, and using the systems), to preserve their authority over the systems, and to safeguard their well-being
Many organizations—35%— believe they have fully implemented an RAI program
Stage 2
Developing
Stage 3
Advanced
Stage 4
Leading
The four stages of RAI maturity
Stage 1: Lagging
Starting to implement an RAI program, with a focus on data and privacy
Stage 2: Developing
Expanding across
the remaining RAI dimensions and initiating RAI policies
Stage 3: Advanced
Improving data
and privacy, but lagging behind on human-related advances
Stage 4: Leading
Performing at a high level across all RAI dimensions
Accountability
Organizations and individuals designing, developing, deploying, or operating an AI system are responsible for the outcomes as well as for ensuring that they are used appropriately and effectively
Transparency and “explainability”
Organizations and individuals designing, developing, deploying, or operating an AI system must be transparent, explaining, as required, the purpose and goals of the system, how it was developed, and how outcomes are reached
Fairness and equity
AI systems are designed to be inclusive, to identify and mitigate sources of bias, and to promote fair outcomes
Safety, security, and robustness
AI systems are designed to be secure and resilient and to have safeguards that reduce the risk of unintended behaviors and outcomes
Data and privacy governance
AI systems are designed, and policies are in place, to ensure compliance with data privacy laws and to mitigate privacy risks
Social and environmental impact mitigation
AI systems are designed to promote positive, sustainable impact and to avoid creating or perpetuating adverse effects on society and the environment
Human plus AI
AI systems are designed to empower people (those who are developing, deploying, and using the systems), to preserve their authority over the systems, and to safeguard their well-being
Relative change in an organization’s focus
Stage 1 to stage 2
Stage 2 to stage 3
Human plus AI
Social and environmental impact mitigation
Data and privacy governance
Safety, security, and robustness
Fairness and equity
Transparency and “explainability”
Accountability
Human plus AI
Social and environmental impact mitigation
Data and privacy governance
Safety, security, and robustness
Fairness and equity
Transparency and “explainability”
Accountability
Human plus AI
Social and environmental impact mitigation
Data and privacy governance
Safety, security, and robustness
Fairness and equity
Transparency and “explainability”
Accountability