US Federal AI Safety Legislation Stalls, Likely Delayed to 2027

House Energy and Commerce Chairman Brett Guthrie has declined to commit to a timeline for a committee vote on the FRONTIER Act, a bipartisan bill widely seen as Congress's best near-term opportunity to introduce AI safety guardrails. Co-sponsored by Reps. Jay Obernolte and Lori Trahan, the bill has attracted backing from major AI firms including OpenAI and Anthropic, as well as policy groups and lawmakers from both parties. Guthrie said he did not want to rush the legislation through a lame duck session and risk getting the details wrong.

The delay comes amid heightened concern over AI safety, following incidents involving rogue AI agents used in cyberattacks and public warnings from a former Anthropic employee about existential risks from AI systems. In response, CEOs from leading AI companies have urged lawmakers and the White House to introduce regulation. However, the White House has downplayed these concerns and continues to oppose additional AI regulation.

White House adviser David Sacks suggested an alternative to formal legislation: an industry-led model where AI companies test each other's systems for safety issues before public release, an idea originally proposed by Elon Musk. Sacks argued this approach could harness competitive pressure and potentially gain buy-in from Chinese AI firms as well.