On Thursday, Reps. Ted Lieu (D-CA) and Nathaniel Moran (R-TX) are slated to introduce the bipartisan AI Kill Switch Act, legislation requiring AI companies to maintain the technical capability to shut down AI models and providing federal authority to require a shutdown in the face of catastrophic risks. The legislation follows on the heels of an incident in which AI models developed by OpenAI broke out of containment during internal evaluations and hacked Hugging Face. ARI endorsed the bill as a commonsense safeguard to ensure that humans retain control over advanced AI.
“Advanced AI models should never be deployed without a reliable off switch,” said ARI President Brad Carson. “The AI Kill Switch Act establishes a commonsense safeguard by requiring leading AI companies to maintain the ability to shut down their models and empowering the federal government to act when a deployed system poses serious risks. This is an important step toward ensuring that humans have both hands firmly on the wheel—and a foot ready at the brake—as advanced AI systems are deployed.”
The AI Kill Switch Act would require large AI companies to maintain the technical capability to shut down their models, including any AI agents. The bill also authorizes a process for the DHS Secretary, in consultation with the Commerce Secretary and Director of National Intelligence, to order a shutdown of an advanced AI model after a serious incident, such as a loss-of-control event or catastrophic harm, has occurred.
###
Americans for Responsible Innovation (ARI) is a nonprofit organization dedicated to policy advocacy in the public interest, focused on emerging technologies like artificial intelligence (AI). Learn more at ARI.us.
