A bipartisan bill introduced in the U.S. proposes the creation of "kill switches" for the most powerful AI models, allowing the Department of Homeland Security (DHS) to order throttling or complete shutdowns in cases of catastrophic harm. The AI Kill Switch Act targets models costing over $100 million to train and imposes daily fines of up to $20 million for non-compliance. This legislation is partly inspired by recent incidents involving OpenAI's GPT-5.6 Sol and Anthropic's Mythos 5, though it exempts issues arising during structured testing like red-teaming. AI
IMPACT Could lead to increased government oversight and control over advanced AI development and deployment.
RANK_REASON Bipartisan bill proposing significant regulatory action for AI models. [lever_c_demoted from significant: ic=1 ai=0.7]
- AI Kill Switch Act
- Anthropic
- Director of National Intelligence
- GPT-5.6 "Sol"
- Mythos 5
- Nathaniel Moran
- OpenAI
- Ted Lieu
- United States Department of Commerce
- United States Department of Homeland Security
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →