Google DeepMind CEO Demis Hassabis calls for a U.S.-led standards body to oversee advanced artificial intelligence development. The proposed agency follows the Financial Industry Regulatory Authority (FINRA) model as an industry-funded but government-accountable organization.
Technical experts would staff the body to establish safety standards and evaluate frontier-class AI models. AI labs would voluntarily submit their most powerful systems for a 30-day review period before public release.
Hassabis argues this systematic approach is necessary to manage escalating cybersecurity and biosecurity risks. He believes Artificial General Intelligence (AGI) could be achieved within several years.