Google, OpenAI, and Anthropic Plan to Launch Their Own AI Safety Standards Body

Google, OpenAI, and Anthropic are moving forward with plans to create their own AI safety standards body, aiming to launch it by the end of this year or early next year. The group is tentatively named the Standards Authority for Frontier AI (SAFA). It would be an industry self-regulatory organization for frontier model developers, operating without government oversight.

SAFA is expected to take on several core functions, including supporting third-party testing of models before they are deployed, setting rules for how AI developers report safety and security incidents, and turning voluntary lab commitments into concrete standards. The working group is also discussing whether SAFA should directly handle safety and capability testing.

Currently, such testing is done by the U.S. Commerce Department's AI standards center, and some members worry it lacks resources. Some have suggested SAFA add extra tests and evaluations on top of the center's work, but it is still unclear whether their roles would overlap. Executives from the three companies have also publicly called for unified safety standards and more international coordination.