What happened
On 29 September 2026, Senate Commerce Committee Chair Ted Cruz (R-TX) blocked a unanimous-consent bid by Sens. Mark Warner, Brian Schatz and Andy Kim to pass the Artificial Intelligence Risk Management and Security Act of 2026. The bill would have created a permanent AI safety board in the Commerce Department to develop frontier-model evaluation standards, require developers to give the board model access at least 45 days before public release, and impose fines for non-compliance. Cruz objected to what he called unbounded executive-branch discretion over private AI companies.
Why it matters
The blocked unanimous-consent vote is the clearest near-term measure of the binding-federal-oversight debate: a proposed mandatory pre-deployment review regime with 45-day model access could not clear even a fast-track procedural hurdle, cementing that no binding federal frontier-AI safety law is imminent this session — reinforcing that compliance pressure for now comes from the FTC/state enforcement front (see FTC probe and Bonta subpoena).
Action needed
AI labs should treat mandatory 45-day pre-deployment federal review as not imminent for now, but should monitor reintroduction; the sponsors intend to keep pressing the bill.