HR 8819: Federal Artificial Intelligence Risk Management Act of 2026

HR 8819 in plain English: This bill is early in the legislative process and detailed text is not yet available. Sponsor: Rep. Lieu, Ted [D-CA-36] (D) · Status: Referred to the House Committee on Science, Space, and Technology.

Stated purpose

This bill requires federal agencies to follow the Artificial Intelligence Risk Management Framework developed by the National Institute of Standards and Technology (NIST) when using or procuring AI systems. It directs NIST to develop formal standards, guidelines, and minimum requirements for how agencies manage risks from AI.

Arguments supporters make

• Federal agencies already use powerful AI in ways that affect people's lives, and requiring a common risk management standard helps ensure those systems are safer and more accountable.
• NIST's framework was developed with broad input from industry and experts, so building federal rules around it avoids reinventing the wheel and aligns government with established best practices.
• Requiring agencies to label and track AI-generated content helps the public know when they are dealing with synthetic material produced by the government.

Arguments opponents make

• Mandating a specific framework across all agencies could create rigid compliance bureaucracy that slows down beneficial AI adoption without meaningfully improving safety outcomes.
• NIST standards are voluntary by design; turning them into binding federal requirements may not be flexible enough to keep pace with how quickly AI technology changes.
• Exempting national security systems from these requirements creates an uneven standard, leaving some of the most consequential government AI uses outside the accountability framework this bill creates.

Tradeoffs

Establishing consistent, mandatory AI risk standards across the federal government may improve accountability and public trust, but could also increase compliance costs and reduce agencies' flexibility to adopt new AI tools quickly. The bill also draws a line between civilian and national security AI systems, applying stricter oversight to one while leaving the other largely unaffected.

Current status in Congress: In committee.