Trump Rejects AI Safety Regulation Amid Congressional Push for Guardrails

Congress returned Monday to a compressed schedule and an AI safety debate that has divided leaders over whether to legislate now or defer to the industry.

• The House is set to adjourn by Sept. 17; the Senate eyes leaving by Oct. 2.
• 27-year-old former Anthropic researcher Jacob Coxon warned publicly that AI could kill all humans by decade's end.
• Anthropic CEO Dario Amodei said he agrees with Coxon's assessment 'much more than I disagree with him.'
• OpenAI CEO Sam Altman and SpaceX CEO Elon Musk backed Amodei's call for a global development slowdown.

Why it matters: The compressed pre-midterm calendar means any legislation would need to move at unusual speed — or wait until after November.

• Senate Minority Leader Chuck Schumer called for an all-senators classified briefing on the administration's AI plans.
• Senators have not yet agreed on provisions holding AI labs legally accountable, leaving no ready bill to pass.

How 76 sources split on this story

Where they split: The central dispute is whether Congress should legislate AI safeguards immediately or let companies and the White House lead while avoiding steps that could slow the U.S. advantage over China.

What’s next: House leaders plan a floor vote on the Ratepayer Protection Act, targeting data center energy costs.

• Senate Commerce may hold an AI safety bill markup this month; a broad bipartisan draft remains unfinished.
• A bipartisan 'pro-human' AI forum featuring Sen. Bernie Sanders and former Trump aide Steve Bannon is set for Sept. 15.
• Will Senate Commerce finalize a bipartisan AI safety bill before the October recess?
• Could Johnson call the House back if Senate reconciliation moves, and would AI legislation attach to that?
• How much political weight will Coxon's warning carry once details of his claims face independent scrutiny?