DeepMind CEO calls for an independent standards body to regulate frontier AI
DeepMind CEO Demis Hassabis proposed an independent regulatory body modeled after FINRA to oversee and test frontier AI models before their release.Summary
Google DeepMind CEO Demis Hassabis has proposed the creation of an independent standards body to oversee the release of frontier AI models. Drawing inspiration from the Financial Industry Regulatory Authority (FINRA), this organization would conduct technical assessments and develop best practices. Initially operating on a voluntary basis, the body could eventually transition to a mandatory requirement for US market deployment.
The initiative seeks to improve upon current ad hoc government reviews, which have faced criticism for lacking technical depth. Hassabis suggests the body should be government-backed but industry-funded and staffed by technical experts and safety advocates. This approach aims to address security vulnerabilities while fostering innovation, even as other policymakers remain skeptical about establishing formal executive oversight for artificial intelligence.
(Source：TechCrunch)