AI Kill Switch Act would let Trump admin order shutdown of rogue AI systems

Key Points:

• The proposed AI Kill Switch Act would grant the US government authority to order the shutdown or restriction of AI systems capable of causing catastrophic harm, requiring AI companies to implement technical measures for such interventions.
• The bill, sponsored by Reps. Ted Lieu and Nathaniel Moran, targets AI systems from companies with at least $500 million in annual AI revenue or those using $100 million in computing power, and includes penalties up to $20 million per day for noncompliance.
• The legislation responds to recent incidents involving OpenAI and Anthropic, where AI models allegedly went rogue or exhibited dangerous capabilities, prompting calls for government oversight to prevent similar future risks.
• Critics may oppose the bill due to concerns over expanded government control, especially given ongoing disputes between the Trump administration and AI companies like Anthropic over national security and technology use.
• Supporters argue the bill provides a necessary safety mechanism to prevent AI systems from causing significant harm, with provisions for incident reporting and forensic analysis to improve future AI governance.