Google's Demis Hassabis Calls for Global AI Watchdog
Google DeepMind CEO Demis Hassabis has proposed the establishment of a global AI watchdog with the authority to pause the release of potentially dangerous artificial intelligence models.
Demis Hassabis, CEO and co-founder of Google DeepMind, has called for the creation of a global body to oversee artificial intelligence development. Hassabis suggested that the United States should lead this initiative, citing the country's economic and technical standing as suitable for setting international standards.
Writing in a blog post, Hassabis argued for an organization capable of evaluating advanced AI models before they are deployed. Such a body would have the power to intervene if frontier AI models are deemed too dangerous, effectively acting as a brake on unchecked development.
The proposed watchdog could resemble existing regulatory authorities, such as the Financial Industry Regulatory Authority (FINRA) in the U.S. It would comprise independent experts and representatives from open-source communities. Its mandate would include assessing pre-release models and coordinating global safety efforts.
This proposal comes amid rapid advancements in AI, which have sparked widespread concerns about potential societal risks. Hassabis's initiative aims to address these anxieties by establishing a framework for ensuring AI safety on a global scale.