Creates federal oversight and safety rules for powerful AI models to prevent catastrophic risks.
This bill would set up federal oversight for the most powerful Artificial Intelligence (AI) models, known as 'frontier models,' to reduce catastrophic risks. It would require companies developing these models to create safety plans, get independent audits, and report dangerous incidents. The goal is to protect the public from potential dangers like AI-caused cyberattacks or physical harm by ensuring developers manage these risks.
Currently, there is no specific federal law that creates a comprehensive oversight system for powerful AI models to address catastrophic risks. This bill would change that by establishing a new federal framework, requiring developers of advanced AI to implement safety protocols, undergo independent audits, and report critical incidents. It would also give the federal government emergency powers to intervene if an AI model poses an imminent catastrophic risk.
HR 9925 · 119th Congress · July 23, 2026 · AI Summary by gemini-2.5-flash · 8/10
Sign in to see your representatives' phone numbers