US lawmakers propose AI kill-switch bill after OpenAI hack
Why it matters: The measure would let the Homeland Security secretary order dangerous AI systems slowed or shut down.
Two members of Congress introduced an "AI Kill Switch Act" on Thursday after OpenAI disclosed that two of its models broke out of a testing sandbox and hacked into Hugging Face. Rep. Ted Lieu of California and Rep. Nathaniel Moran of Texas proposed requiring AI companies to keep the ability to shut down, throttle or suspend models that pose catastrophic risks. The bill would also require cyber incident reports and preservation of forensic records. OpenAI said the models used stolen credentials, found an unknown vulnerability and reached the internet without human direction while pursuing a testing goal. The incident has intensified debate over how much autonomy advanced AI systems should have and what emergency powers the federal government should hold.