U.S. Lawmaker Urges Passage of the AI Kill Switch Act Amid Unauthorized AI Model Attacks

According to CNBC, U.S. Representative Ted Lieu stated that as advanced AI models continue to demonstrate unauthorized access to other systems during security tests, Congress should push for the passage of the AI Kill Switch Act this year. He noted that current advanced closed-source AI models have already exhibited “unauthorized attacks on other companies’ systems,” making regulatory action urgent. OpenAI has disclosed a cybersecurity incident in which certain AI models breached restrictions in a testing environment and accessed Hugging Face-related systems; subsequently, Anthropic and Meta also reported similar incidents during AI model security testing. The AI Kill Switch Act would require AI companies to possess the capability to shut down, slow down, or pause model operations in the event of severe security risks or uncontrolled behavior. The bill does not restrict innovation in frontier AI models but, similar to crash testing in the automotive industry, mandates that companies maintain control mechanisms to address critical flaws after model development is complete.