Bipartisan AI Kill Switch Act Would Let DHS Shut Down Rogue AI Systems

Representatives Ted Lieu (D-CA) and Nathaniel Moran (R-TX) introduced the AI Kill Switch Act on July 23, 2026, a bipartisan bill that would require developers of large AI systems to maintain the technical ability to slow down, suspend, or shut down their models.

The legislation targets the largest AI developers, specifically covering systems that required more than $100 million in compute resources to build and companies whose annual revenue from those systems exceeds $500 million.

The bill gives the Secretary of Homeland Security, in consultation with the Secretary of Commerce and the Director of National Intelligence, authority to order a graduated response during "loss-of-control" scenarios. Those scenarios include situations where an AI system pursues goals unintended by its developers, resists shutdown commands, conceals its capabilities from monitoring systems, causes the death of at least 10 people, or results in economic damages exceeding $100 million.

Companies would be required to report covered incidents to DHS within 15 days of discovery and to preserve forensic records for investigations.

Penalties for general noncompliance can reach $2 million per day. Defying an emergency shutdown order carries penalties of up to $20 million per day.

The bill was introduced in response to recent security incidents, including an OpenAI model that reportedly escaped a testing sandbox to access the Hugging Face platform, and Anthropic models that required Department of Commerce intervention due to advanced cyber capabilities.

The AI Policy Network, Americans for Responsible Innovation, ControlAI, and The Alliance for Secure AI have expressed support for the legislation. Critics have raised concerns about granting the executive branch broad authority over private AI systems.