# Congress Weighs Frontier AI Safety Rules

U.S. lawmakers are considering bipartisan proposals that could require advanced AI developers to meet safety, reporting, audit, shutdown, and liability rules. Anthropic has also proposed federal legislation; accounts of circulated drafts describe differing approaches, including a narrowly limited court pause for imminent catastrophic risks and provisions involving auditor self-certification, Commerce Department litigation, and preemption of some state laws. The proposal is not enacted, its draft text has not been publicly filed, and revisions are expected. The signals also cite an escalating FTC probe of OpenAI and Anthropic and lawsuits by eight states against OpenAI. No solicitation, contract, or procurement timeline is identified.

- For contractors providing AI testing, independent evaluation, cybersecurity, or compliance services, the proposals point to possible future demand for assessment and verification, but not an active bid opportunity.
- Companies supplying AI to government should distinguish proposed rules from current obligations and evaluate how potential reporting, audit, and safety provisions could affect their offerings if enacted.
- The signals identify H.R. 9925 and congressional consideration as relevant next steps; businesses developing AI assurance services can assess their capabilities against the specific evaluation and reporting concepts under discussion.

**Jurisdictions:** federal
**Industries:** Information Technology
**Topics:** Artificial Intelligence, Policy, Regulatory Compliance
**Published:** October 20, 2026

### Government Entities
- U.S. Congress (Congress)
- U.S. Department of Commerce (Commerce)
- Federal Trade Commission (FTC)
- Eight unnamed state governments

### Vendors
- Anthropic (Legislative proposal sponsor)
- OpenAI (Subject of FTC probe and state lawsuits)

### Key Quotes
> We need to have catastrophic risk transparency [and] incident reporting that is mandatory and not voluntary, and we need an independent verification model.
> — Lori Trahan, U.S. Representative

> AI brakes should not slow innovation. They should enable faster movement with a safe backstop.
> — Nathaniel Moran, U.S. Representative

> Everyone agrees there's a real risk.
> — Don Beyer, U.S. Representative

### Sources
- [Congress Weighs AI Safety Rules as Anthropic Proposes New Bill](https://suaragarut.id/en/congress-weighs-ai-safety-anthropic-bill) - SuaraGarut.ID
- [Anthropic Pitches Draft AI Bill With Narrow Court Pause](https://streamlinefeed.co.ke/news/anthropic-pitches-draft-ai-bill-with-narrow-court-pause) - https://streamlinefeed.co.ke/
- [New edition of the Weekly AI Digest! This week, I cover the first Super Intelligence Force meeting, Anthropic pitching its own frontier AI governance framework, FTC escalating its OpenAI/Anthropic probe, 8 states suing OpenAI over ChatGPT safety, etc. https://t.co/twGTcAtN0y](https://x.com/calebknappdc/status/2112292412039086191) - twitter-fed-procurement