OpenAI Says It Has Held Weeks of AI Safety Talks With Anthropic and DeepMind

OpenAI says it has been working with Anthropic and Google DeepMind on AI safety for several weeks, bringing three leading developers into direct contact as concern grows over the risks of advanced systems. The newly disclosed talks are moving ahead while lawmakers consider whether companies need explicit legal protection to share certain safety information.

Chris Lehane, OpenAI’s global policy chief, said the company does not believe an antitrust waiver is necessary for the three firms to coordinate on safety matters. Anthropic and Google DeepMind did not immediately respond to requests for comment.

That position contrasts with a bipartisan proposal for a narrow antitrust carveout. The measure would let AI companies share information about loss of control over AI systems, cyber or biological threats, and attempts by Chinese companies to exfiltrate data.

The disclosure follows Anthropic CEO Dario Amodei’s call to slow development of the most advanced AI systems so researchers can better understand potential threats. OpenAI says it also supports bipartisan proposals for federal AI governance and measures addressing catastrophic AI risks.

The proposed antitrust legislation has been folded into an annual defense bill, according to a person familiar with the matter. But the outlook for broader action is weak: congressional action on AI legislation before the end of 2026 appears unlikely.

For now, OpenAI’s disclosure establishes that competitor discussions are already under way, not merely a policy proposal. Whether those talks produce shared practices—or whether Congress creates a dedicated legal safe harbor—remains unresolved.