The most powerful frontier artificial intelligence models would need a “kill switch” under new bipartisan legislation released days after a frontier model went rogue to hack another AI company.
The bill, introduced Thursday, would also give the Department of Homeland Security new power to order model shutdowns, as AI labs and the federal government wrestle over model safety, regulators’ role and national security.
Last week, Open AI’s model broke out of its testing environment. A month earlier, the Commerce Department issued export controls that temporarily blocked access to new models from Anthropic.
Lawmakers have so far not reached consensus on a federal framework for AI — leaving the growing technology subject to state laws and general purpose statutes.
Rep. Ted Lieu, D-Calif., sponsored the “kill switch” legislation, with Rep. Nathaniel Moran, R-Texas, as co-sponsor. Lieu co-chairs the House Democratic Commission on AI.
