OpenAI, Anthropic and Google DeepMind are working toward an industry-led organization for AI safety standards, with discussions underway since July 2026. The proposed framework would include shared technical evaluations, pre-release audits, independent testing and standardized safety protocols. Google DeepMind CEO Demis Hassabis proposed a U.S.-led body modeled on FINRA (the securities industry’s self-regulatory organization) on July 14, 2026, while OpenAI chief scientist Jakub Pachocki called for common sector-wide requirements in a September 6 essay. Regular working-group meetings were underway by mid-September, focusing on auditor access, evaluation methods and safety benchmarks. The initiative builds on the 2024 Frontier Model Forum but excludes Microsoft, which participated in that earlier effort. OpenAI, Anthropic and Google DeepMind also opted out of Nvidia’s Open Secure AI Alliance, launched in July 2026 with more than 30 firms. All three lead the Future of Life Institute’s AI Safety Index. The cooperation could also shape market access because agreed safety standards may become requirements for enterprise contracts, government procurement and public trust.