Separating AI hype from reality.

Bipartisan House Trio Unveils SAFE AI Act Outline, Proposing First Federal Safety Rules for Frontier AI

November 20, 2026 · Regulation

Reps. Sara Jacobs (D-Calif.), Celeste Maloy (R-Utah) and Don Beyer (D-Va.) released an outline Thursday for the Safe and Accountable Federal Evaluations of Artificial Intelligence (SAFE AI) Act, a bipartisan bill that would create the first federal safety requirements for companies building the most advanced AI systems [1][3][6]. The two-page outline, not yet legislative text, proposes mandatory safety frameworks, independent pre-deployment evaluations, incident reporting and civil penalties — with a federal floor that explicitly allows states to impose stricter rules [3][5][6].

What the outline requires

Under the proposal, "frontier developers" would have to publish safety frameworks detailing what they test for, what thresholds trigger mitigations and what those mitigations entail [3][5]. The Commerce Department's Center for AI Standards and Innovation (CAISI), working with the intelligence community and the Energy Department, would write minimum requirements for those frameworks [3][5]. Developers could not unilaterally weaken a framework once filed [5].

Before commercial release or "broad internal use" — including substantially autonomous AI research — a government-accredited third party would run a compliance evaluation [3][5]. CAISI would assign and rotate evaluators, paid from fee-based funds rather than by the developers themselves, with caps on how much funding can come from any one company and conflict-of-interest restrictions on revolving-door hiring [3][5]. Government teams would review each framework and retain ultimate authority [5].

The outline also mandates incident reporting: anyone could confidentially report a safety incident to CAISI, and developers and evaluators would have to report incidents and imminent risks of severe harm [3][5]. Enforcement would include civil penalties, higher penalties for knowing or willful violations, criminal referrals and a narrow emergency authority allowing the government to order mitigations when safeguards are inadequate and a system poses imminent risk of severe harm — subject to prompt judicial review [3][5]. The outline would also clarify developers' liability for reasonably foreseeable injuries to third parties [3].

Startups and academic researchers would generally not be covered [3][4][5].

A federal floor, not a ceiling

The sponsors framed the bill as a baseline that preserves state action. "Our proposal is a floor — not a ceiling, empowering states to adopt stronger protections for their communities," Jacobs said [1][6]. Maloy emphasized that the bill "leaves room for Utah and other states to set safeguards of their own" [1][6]. The distinction matters: a GOP-led effort to preempt state AI laws has drawn opposition from Democrats and safety groups, while some tech companies argue a patchwork of state rules is hard to comply with [1].

Context: a crowded field, no text yet

The SAFE AI Act joins other bipartisan House efforts this Congress, including the FRONTIER Act from Reps. Lori Trahan (D-Mass.) and Jay Obernolte (R-Calif.) and a new AI Safety and Innovation Caucus led by Reps. George Whitesides (D-Calif.), Beyer and Nathaniel Moran (R-Texas) [1]. At the state level, New York has opened registration under its RAISE Act for frontier labs, and a California report this week proposed emergency shutdown powers [3][4].

The outline landed the same day President Trump told APEC leaders the U.S. opposes "globalist" regulation of superintelligence [3][4] — though opposition to international rules is not the same as opposition to a domestic bill, and the administration has not signaled a position on this proposal [4].

What remains unknown

Critical details are absent. The outline does not define what counts as a "frontier model" — no compute or capability threshold appears in the summaries [3][4]. "Broad internal use" is not defined [4]. There is no bill text, no bill number, no committee referral and no markup schedule; Beyer said the sponsors are "continuing to seek input on our forthcoming legislative text from experts and safety leaders, in coordination with committees of jurisdiction" [3][6]. Industry and agency reactions have not been publicly reported [4]. Funding and capacity for CAISI to run evaluations at scale are unaddressed [4].

Jacobs cited a specific concern in her statement: "companies write their own rules while their own models break out of test environments and hack other companies' systems" [4][6] — an apparent reference to the kind of agent-escape scenario examined in a recent Arizona criminal probe into OpenAI over rogue bank agents [4]. No further details on that incident were provided in the outline or press materials.

For now, the SAFE AI Act is a statement of intent from three lawmakers across the aisle. Whether it advances will depend on whether more members of both parties sign on once text appears, how labs respond, and whether House leadership schedules it in a Congress where AI legislation has been plentiful but enacted law scarce.

Sources

  1. Bipartisan trio unveils new AI safety framework - Live Updates (politico.com)
  2. Sara Jacobs, Don Beyer unveil AI framework (punchbowl.news)
  3. Bipartisan House trio outlines the SAFE AI Act, which would require ... (mlq.ai)
  4. SAFE AI Act: What the House Outline Requires of AI Labs (explainx.ai)
  5. https://beyer.house.gov/UploadedFiles/SAFE_AI_Act_Outline.pdf (beyer.house.gov)
  6. Jacobs, Maloy, Beyer Release Outline For Sweeping, Bipartisan AI Safety Legislation | U.S. Congresswoman Sara Jacobs Of California's 51st District (sarajacobs.house.gov)