OpenAI, Google and Anthropic are jointly developing a new industry body focused on safety standards for frontier AI systems. According to a report from The Information on September 24, the organization is tentatively called the Standards Authority for Frontier AI, or SAFA.

The three labs aim to launch it by the end of 2026 or early 2027. The body would support third-party testing of models before deployment, set rules for how developers report safety and security incidents, define voluntary safety commitments, and establish qualifications for independent auditors.

Earlier efforts to form a public-private partnership stalled under the current administration. The new plan is positioned as industry self-regulation without direct government oversight. The existing Frontier Model Forum, created in 2023 by the same companies plus Microsoft, remains active and is expected to coordinate with any new entity.

Key takeaways

What SAFA’s mandate would actually cover

The proposal emphasizes practical mechanisms over broad principles. Third-party organizations would test models before they reach users. Labs would follow defined reporting protocols for safety and security incidents. Independent auditors would need clear qualifications. Critics have already flagged the risk that the three largest closed-model developers could use the body to disadvantage open-source competitors.

Primary reporting: The Information. Related coverage also appeared on PYMNTS and industry briefings. For broader context on how AI tools shape creator workflows, see How AI is Reshaping the Content Creator Stack in 2026.