HR 9333: AI Flaw Reporting and Security Enhancement Act

HR 9333 in plain English: This bill is early in the legislative process and detailed text is not yet available. Sponsor: Rep. Ross, Deborah K. [D-NC-2] (D) · Status: Ordered to be Reported in the Nature of a Substitute by the Yeas and Nays: 35 - 0.

Stated purpose

This bill directs the National Institute of Standards and Technology (NIST) to create a voluntary program for reporting, collecting, and tracking flaws in artificial intelligence systems, and to build tools and a national database to detect and monitor those flaws.

Arguments supporters make

• A common language and national database for AI flaws would make it easier to find, track, and fix problems before they cause serious harm to people.
• The program is voluntary, so it encourages cooperation from industry and researchers without imposing new government mandates on businesses.
• Bringing together companies, academics, and government to set shared standards creates a more coordinated and effective response to AI safety and security risks.

Arguments opponents make

• Because participation is voluntary, companies with the most serious AI flaws may simply choose not to report them, leaving the database incomplete and the public underprotected.
• Creating definitions, taxonomies, and standards through a large multi-stakeholder process could take years, during which AI risks continue to grow without meaningful oversight.
• A national database of known AI flaws, if not carefully secured, could itself become a target that bad actors exploit to find and use those vulnerabilities.

Tradeoffs

Keeping the program voluntary makes it easier for industry to participate and share information, but may result in incomplete reporting of the most serious flaws. Building shared definitions and infrastructure takes time and broad coordination, which could slow action while AI systems are already widely deployed.

Current status in Congress: In committee.