AI FINRA: US Considers Watchdog for AI Models Before Release

The United States is considering creating an independent watchdog to vet advanced AI models before public release. This proposed regulator would be modeled on FINRA, the body overseeing America's brokers. The proposal suggests an independent AI regulatory agency reporting to the Securities and Exchange Commission. Frontier AI labs would submit their most capable models for a 30-day review before release. Experts would screen for cyber weapons capability, biological weapons capability, and deception. Participation would initially be voluntary, with the potential to become mandatory. This push comes after Silicon Valley leaders complained about unpredictable government actions regarding cutting-edge AI systems. Industry leaders like Demis Hassabis, Satya Nadella, Sam Altman, and Elon Musk have expressed support for similar policy recommendations. The FINRA model means the AI labs themselves would likely fund and staff their own watchdog, with the SEC providing oversight. This approach leverages industry expertise and could move faster than Congress. However, critics point out that a voluntary review of chosen models may not guarantee safety, and questions remain about funding and the SEC's role. This development could significantly impact how new AI technologies are introduced to the public.