Learning From Other Models
The RFC asked what accountability policies adopted in other domains might be useful precedents for AI accountability policy. Commenters addressed this question in detail.
-
Financial Assurance
The assurance system for financial accounting is an obvious referent for AI assurance. Some existing financial sector laws may be directly applicable to AI. Otherwise, they may still furnish useful analogies. In other words, as one commenter stated, “the established financial reporting ecosystem provides a valuable skeleton and helpful scaffolding for the key components needed to establish an AI accountability framework.”
-
Human Rights and ESG Assessments
The private sector continues to refine and seek ESG framework standardization for evaluations. What the ESG assurance experience may teach is that multi-factored evaluations using a variety of standards may not immediately yield comparable or actionable results.
-
Food and Drug Regulation
Another potentially useful accountability model suggested by commenters can be found in health-related regulatory frameworks such as the FDA’s. FDA regulates some AI systems as medical devices.
-
Cybersecurity and Privacy Accountability Mechanisms
We recommend that future federal AI policymaking not lean entirely on purely voluntary best practices. Rather, some AI accountability measures should be required, pegged to risk.