Cruz blocks Senate Democrats’ effort to pass AI safety bill
The Texas senator objected to a proposed AI Safety Board, calling it an overreach of executive power over private companies.
Senator Ted Cruz blocked a Democratic push to advance sweeping AI safety legislation on September 29, killing the bill’s path through unanimous consent on the Senate floor. The Artificial Intelligence Risk Management and Security Act of 2026, sponsored by Senators Mark Warner, Brian Schatz, and Andy Kim, would have created a permanent AI Safety Board inside the Department of Commerce with real enforcement teeth.
Cruz, who chairs the Senate Commerce Committee, objected on the grounds that the bill would hand the federal government what he described as “unbound discretion” in regulating private AI companies.
What the bill would have done
The bill would have established an AI Safety Board within the Commerce Department, empowered to enforce compliance standards for frontier AI models. The most consequential provision: developers would have been required to grant the board access to their models a full 45 days before any public release.
The bill emerged partly in response to warnings from AI researchers and a string of security breaches at major labs. Warner, Schatz, and Kim framed it as a necessary guardrail for technologies that are evolving faster than the government’s ability to understand them, let alone regulate them.
The partisan fault lines
Senator Maria Cantwell has been among the most vocal advocates for stronger federal testing mandates, directly countering proposals that would let companies self-certify their safety practices.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
Cruz didn’t simply reject regulation outright. He proposed alternative legislative routes that, in his framing, would avoid compromising corporate interests while still addressing safety concerns.
Cruz, Senate Majority Leader John Thune, and Senator Amy Klobuchar had been engaged in negotiations aimed at addressing catastrophic risks from AI. Those discussions have stalled, stuck on the fundamental question of how much authority the government should have over the companies building these systems.
Why the timing matters
With November midterms approaching, the window for meaningful AI legislation is narrowing fast. Lawmakers on both sides of the aisle have acknowledged that any substantial action might have to wait for a lame-duck session after the elections.