Mapping AI Bill of Right’s requirements on the HAIP Guides

AI Bill of Rights Principles	Requirement	HAIP Guides
Safe and effective systems	Multi-stakeholder involvement Identify risks Pre-deployment testing, risk mitigation Monitoring Domain specific use Public documentation of safety	Identify problems across the organization Identify potential downstream impacts Evaluate internal resources for adoption Define the role of AI Locally validate prior to integration Identify and mitigate risks Disseminate information to end users Prevent inappropriate use of the AI product Monitor AI performance Audit AI solutions and work environment Identify potential risks
Algorithmic discrimination preventions	Ensure systems are equitable Assess for equity issues Clear organizational oversight Continuous testing Transparent reporting and impact assessments	Prioritize problems Identify potential downstream impacts Determine suitability of technical approaches Assess quality of external AI products Define performance targets Locally validate prior to integration Identify and mitigate risks Determine if AI should be integrated Monitor AI performance Monitor work environment Audit AI solutions and work environment Identify potential risks Disseminate information about updates to end users
Data privacy	Individual agency in data collection and use Transparent communication about user data Follow ethical review Document data decisions	Assess viability of buying or building Assess legal risks Define the role of AI Define performance targets Disseminate information to clinical end users Disseminate information about updates to end users
Notice and explanation	Accessible documentation Communicate outcomes and updates Notify affected parties Public information accessible	Audit investment decisions Define performance targets Disseminate information to clinical end users Audit AI solutions and work environment Minimize disruptions from decommissioning Disseminate information about updates to end users
Human alternatives, considerations, and fallbacks	Ability to opt out of technological systems   Available human alternatives   Tailor systems to purpose   Maintain system oversight   Public document on human governance	Define the role of AI Design and test workflow for clinicians Prevent inappropriate use of AI Monitor AI performance Monitor work environment Audit AI solutions and work environment Determine if updating or decommissioning is necessary Minimize disruptions from decommissioning