On December 5, the Center for American Progress submitted comments on regulations.gov under the Office of Management and Budget’s (OMB) request for public comment on its draft guidance “Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence.”
You can view CAP’s comments on the draft OMB artificial intelligence (AI) guidance here, and they are available in the public record here.
The comments note:
The draft OMB AI guidance puts principles from the 2022 White House Blueprint for an AI Bill of Rights into effect for the federal government’s use of AI. The White House AI Bill of Rights “identified five principles that should guide the design, use, and deployment of automated systems to protect the American public in the age of artificial intelligence” along with examples of how to implement those principles in practice. The draft OMB AI guidance integrates aspects of the AI Bill of Rights in its guidance to federal agencies within the limits of executive authority. … Particularly worthy of praise is the draft guidance’s definition of safety-impacting and rights-impacting AI and purposes that are presumed to be safety-impacting and rights-impacting. These thoughtful definitions are an important step forward in outlining the conditions in which the use of AI requires extra attention and caution.
The comments go on to outline 13 specific concerns about the draft OMB AI guidance, including where clarifications or additions are needed around disability and worker rights; transparency for waivers and exemptions to the minimum practices; and the importance of emphasizing the option to not adopt AI for safety-affecting or rights-affecting uses. The comments also answer questions 5, 6, and 8 posed by the OMB in its Federal Register posting.
CAP also authored and submitted separate comments to the OMB on behalf of a broad group of more than 17 civil, tech, environmental, and human rights organizations. Those comments can be viewed here and are available in the public record here.
The author would like to thank Megan Shahi, Hauwa Ahmed, Sydney Bryant, Mia Ives-Rublee, Ben Olinsky, Frank Torres, Richard Anthony, and Dr. Alondra Nelson for their contributions to this comment.