Tech Giants Push Ahead with Independent AI Safety Standards Body
Google, OpenAI, and Anthropic are pushing forward with plans to establish an industry-led AI safety standards body, which would operate independently without government oversight. The group, expected to be called the Standards Authority for Frontier AI (SAFA), aims to launch by late 2026 or early 2027.
SAFA would set standards for testing and reporting AI safety issues, establish guidelines for voluntary safety commitments made by AI labs, and define how independent auditors evaluate AI models and labs. Some members of the working group are debating whether SAFA should conduct additional evaluations alongside the Center for AI Standards and Innovation (CAISI), which some believe lacks sufficient resources to thoroughly test frontier models.
High-profile figures have been considered for leadership roles, including Sriram Krishnan, a former venture capitalist and top AI policy adviser in the Trump administration, and Condoleezza Rice, former U.S. Secretary of State. However, the government's role in overseeing SAFA remains unclear, with some company representatives remaining optimistic about a future relationship.