OpenAI, Anthropic, and Google DeepMind collaborate on AI safety standards body

OpenAI, Anthropic, and Google DeepMind are in ongoing talks to establish an industry-led safety standards body for artificial intelligence, inspired by a proposal from Demis Hassabis. While leaders like Sam Altman and Elon Musk support pacing development to mitigate catastrophic risks, others, including Meta's Mark Zuckerberg and Nvidia's Jensen Huang, oppose such regulation. Meanwhile, OpenAI is actively lobbying U.S. lawmakers, expressing support for independent safety audits under the FRONTIER Act.

AI safety standards body

• ▪United States House Speaker Mike Johnson stated in September 2026 that there is no consensus among competing artificial intelligence companies regarding safety standards and guardrails
• ▪OpenAI, Anthropic, and Google DeepMind have been engaged in discussions since July 2026 to establish an industry-led standards body for artificial intelligence safety
• ▪OpenAI, Anthropic, and Google DeepMind opted out of Nvidia's Open Secure AI Alliance, which launched in July 2026 to focus on open-source artificial intelligence security tools
• ▪The proposed collaboration between OpenAI, Anthropic, and Google DeepMind focuses on shared technical evaluations, pre-release audits of advanced models, independent testing frameworks, and standardized safety protocols
• ▪The discussions between OpenAI, Anthropic, and Google DeepMind regarding an artificial intelligence safety standards body began before the September 2026 resignation of a former Anthropic researcher

Demis Hassabis FINRA proposal

• ▪Google DeepMind CEO Demis Hassabis proposed on July 14, 2026, a United States-led artificial intelligence standards organization modeled after the Financial Industry Regulatory Authority
• ▪Google DeepMind CEO Demis Hassabis proposed that the artificial intelligence standards body be a public-private partnership overseen by the government, funded by the industry, and staffed by independent technical experts
• ▪Following Demis Hassabis's July 2026 proposal, Bloomberg reported that United States Treasury Secretary Scott Bessent was considering a Financial Industry Regulatory Authority-like independent regulatory agency for artificial intelligence reporting to the Securities and Exchange Commission

Dario Amodei slowdown essay

• ▪Anthropic CEO Dario Amodei published an essay on September 12, 2026, calling for artificial intelligence development labs to slow down the pace of leading-edge model upgrades to map and defuse hazards
• ▪Former Anthropic researcher Jacob Coxon resigned in September 2026 and warned that artificial intelligence companies were gambling with lives, a concern echoed by former Google DeepMind researcher Bilal Chughtai

Industry leader support

• ▪OpenAI CEO Sam Altman, Google DeepMind CEO Demis Hassabis, and xAI head Elon Musk supported Anthropic CEO Dario Amodei's call to slow the pace of frontier artificial intelligence development
• ▪United States President Donald Trump dismissed warnings about artificial intelligence risks as a hoax and opposed tighter regulations, stating that any slowdown would benefit China
• ▪Meta CEO Mark Zuckerberg opposed a national artificial intelligence regulator and advised United States President Donald Trump against the idea in a phone call during the summer of 2026
• ▪Nvidia CEO Jensen Huang stated in September 2026 that predictions of artificial intelligence destroying humanity are not grounded in science and called such warnings irresponsible

Antitrust law concerns

• ▪OpenAI CEO Sam Altman and other industry leaders noted that safety coordination talks among rival artificial intelligence firms could risk violating antitrust laws if found to suppress competition
• ▪OpenAI global policy chief Chris Lehane stated on September 15, 2026, that OpenAI, Anthropic, and Google DeepMind do not require an antitrust exemption to collaborate on safety standards

Third-party safety evaluators

• ▪OpenAI CEO Sam Altman stated that OpenAI would join Anthropic in embedding third-party evaluators into the company to monitor artificial intelligence safety
• ▪OpenAI global policy chief Chris Lehane stated on September 15, 2026, that OpenAI supports a provision in the FRONTIER Act forcing frontier labs to allow independent verification organizations to ensure safe model development
• ▪OpenAI chief scientist Jakub Pachocki stated in September 2026 that OpenAI has been discussing concrete safety standards with external organizations and plans to share updates in the coming months

Debatable claims

• ▪AI labs should slow the development of frontier models
• ▪Governments should mandate independent safety audits for advanced AI models
• ▪AI companies should self-regulate safety instead of lobbying for government rules