Proposed Federal AI Legislation and the Role of NIST
AI Auditing Legislation Gains Traction in Congress
Legislation related to artificial intelligence (AI) auditing is gaining momentum in Congress, with several bills proposing different approaches to overseeing AI systems. The Validation and Evaluation for Trustworthy Artificial Intelligence Act (VET AI Act), introduced by Sen. John Hickenlooper, focuses on voluntary guidelines for AI auditing, in contrast to other bills that propose more regulatory authority for overseeing AI.
Currently, over 760 AI-related bills are pending in the United States, with a mix of federal and state-level proposals. The VET AI Act stands out for its emphasis on collaboration and voluntary standards, working with the National Institute of Standards and Technology (NIST) to develop guidelines for AI auditing.
AI auditing and impact assessments are becoming increasingly important tools for evaluating the performance and potential harms of AI systems. While some bills propose mandatory assessments, the VET AI Act takes a more flexible approach, focusing on best practices and certification of third-party evaluators.
The role of the Department of Commerce (DOC) in AI policy is also expanding, with NIST playing a central role in developing AI safety standards. The VET AI Act and other Senate bills envision NIST as a key player in overseeing AI governance, building on existing frameworks like the AI Risk Management Framework.
Recent Supreme Court decisions, such as Loper Bright v. Raimondo and Murthy v. Missouri, could impact the implementation of AI auditing legislation. While the Court’s decisions may influence the regulatory landscape, the focus on voluntary standards and collaborative governance remains a key aspect of the proposed AI legislation.
Overall, the VET AI Act and other AI auditing bills signal a shift towards more collaborative and flexible approaches to AI governance, emphasizing best practices and industry standards over rigid regulatory frameworks. As the debate over AI policy continues, finding the right balance between innovation and oversight will be crucial for the future of AI technology in the United States.