The Threat of Human Extinction Will Get Congress to Act on AI Safety…Right?
As many AI researchers have been increasingly fraught with existential terror about their own creations this summer, their alarm is spreading among policymakers and the media. There are some policy ideas to address the risks: requiring “kill switches” for AI models, setting federal standards for safe research, or even shutting down development of cutting-edge “frontier” […]
As AI researchers have grown increasingly concerned about their own creations this summer, their alarm is reaching policymakers and the media. Various policy proposals exist to mitigate the risks, including requiring "kill switches" for AI models, establishing federal standards for safe research, or even halting development of cutting-edge "frontier" models.
However, enacting a consistent national policy appears unlikely this session, despite calls from AI developers for self-imposed regulations to prevent potential widespread doom. Computer scientist Stephen Casper, who studies AI safety and governance at Harvard Kennedy School, explained that there is uncertainty over whether leading companies could actually shut down their frontier models in an emergency.
Multiple bills have been introduced in Congress to address these concerns. Reps. Nathaniel Moran (R-Texas) and Ted Lieu (D-Calif.) proposed the AI Kill Switch Act, which would require companies to be able to throttle their models and allow top federal officials to order a shutdown if necessary. The FRONTIER Act, spearheaded by Reps.
Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.), would require leading companies to consult third-party experts to ensure adherence to safety protocols. Should these measures fail, independent verifiers would alert the secretary of commerce, who could then shut down frontier model usage. While AI safety policy is less polarized than other political issues, there are still divisions among lawmakers, with some Republicans reluctant to impose regulation.
Congress remains gridlocked, and legislation has not yet been brought to a vote. In the meantime, states such as California, New York, and Illinois have enacted policies requiring developers to publish safety plans and audit their practices. AI companies themselves have published guidelines advocating for third-party inspections of safety practices, with OpenAI and Anthropic emphasizing the importance of government involvement to prevent runaway research.
Despite increased urgency due to recent cybersecurity risks, Congress has yet to overcome partisan divides, with potential breakthroughs grounded in the rapid advancements being made by the most powerful AI models.
Written by urgent.news from Mother Jones's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
