PRIA Logo™
119-hr-10362119th CongressWALLET

Stop Rogue AI Act

Sponsored By: Representative Gottheimer, Josh [D-NJ-5]

Introduced

Summary

Continuous, auditable, cryptographically verifiable discovery and control of AI agents. This bill would require organizations and federal systems to find, verify, monitor, and control AI agents so agencies can audit and govern their behavior. It directs the National Institute of Standards and Technology (NIST) to publish vendor-agnostic standards within 1 year and update them annually.

Show full summary
  • Federal agencies would receive binding procurement and implementation guidance from the Office of Management and Budget (OMB) in coordination with the Cybersecurity and Infrastructure Security Agency (CISA). That guidance must be issued within 180 days after proposed standard revisions and updated each year.
  • Organizations that deploy or interact with AI agents would need continuous, machine-readable inventories, cryptographically verifiable AI identities, and organizational controls to allow or block agent actions. They must evaluate security and reliability before deployment and maintain tamper-evident, standardized logs accessible to authorized parties.
  • NIST would set open, interoperable discovery standards that apply to internal, third-party, and externally hosted agents. Requirements include runtime monitoring and inline detection for prompt injection, data exfiltration, anomalous tool use, behavioral drift, and cryptographic provenance of agent creators/operators.

Personalized for You

How does this bill affect your finances?

Personalize government policy and PRIA will tell you what this bill means for your household, plus every other piece of legislation we track. PRIA reads each provision against your financial profile to show you exactly what matters to your wallet.

Free to download

Bill Overview

Analyzed Economic Effects

2 provisions identified: 0 benefits, 0 costs, 2 mixed.

New AI security standards for organizations

If enacted, NIST would have to publish standards for secure AI agents within 1 year and update them every year. The standards would require continuous, machine-readable inventories of AI agents, cryptographically verifiable agent identity and provenance, runtime monitoring, and tamper-evident logs. The bill would also create official definitions for terms like "AI agent" and "AI agent discovery" used by the standards. These changes would aim to improve security and auditability but would add new technical and compliance work for organizations.

New AI procurement rules for contractors

If enacted, the Federal Acquisition Regulatory Council would have to propose FAR revisions within 18 months after NIST publishes the standards and annually after. OMB, working with CISA, would have to issue agency guidance within 180 days after those proposals and annually. New contracts (not those signed before enactment) would need continuous AI-agent inventories, cryptographic agent identity and provenance, agency access to discovery and controls, and tamper-evident logs. These rules would give agencies more visibility and control but would raise compliance costs for contractors and cloud providers.

Sponsors & CoSponsors

Sponsor

Gottheimer, Josh [D-NJ-5]

NJ • D

Cosponsors

  • Rep. Huffman, Jared [D-CA-2]

    CA • D

    Sponsored 10/15/2026

  • Rep. Lawler, Michael [R-NY-17]

    NY • R

    Sponsored 9/14/2026

Roll Call Votes

No roll call votes available for this bill.

View on Congress.gov
Back to Legislation