AI/TECH
Will a 'FINRA for AI' Actually Keep You Safe?
Google DeepMind wants a financial-style watchdog for AI. But the model it's copying has a messy conflict-of-interest past.
FortuneGoogle DeepMind CEO Demis Hassabis is pushing for a U.S.-led, industry-funded public-private watchdog to screen frontier AI models before deployment, borrowing the framework of financial regulatory body FINRA. However, the proposal is drawing scrutiny because FINRA itself has long faced serious questions regarding conflicts of interest.
- Demis Hassabis proposed a U.S.-led, industry-funded public-private AI standards and testing body.
- The proposed agency would screen frontier artificial intelligence models before U.S. deployment.
- The model is based on FINRA, the self-regulatory body for U.S. financial brokerages.
- FINRA has historically faced heavy criticism over potential conflicts of interest, complicating its appeal as an AI blueprint.
WHY THIS MATTERSWHY THIS MATTERS: Big Tech is trying to design the rules of the road before the government does. If a weak watchdog gets locked in, you get zero real safety guarantees while startups get locked out.