Chris Lehane, the global policy chief at OpenAI, announced on Tuesday that the firm has spent several weeks in discussions with Google and Anthropic regarding industry guidelines. The 3 companies are exploring the creation of a joint standards body focused on the testing and auditing of AI models. Lehane stated that the collaboration on these specific safety matters does not necessitate an antitrust waiver for the firms to coordinate their efforts.

CEO Sam Altman told OpenAI staff that the major labs may be forced to establish this auditing organization independently if U.S. government support is unavailable. This push for external oversight has created internal friction at both OpenAI and Anthropic, where some senior employees worry that allowing third party evaluators deep system access could jeopardize security and leak sensitive technology.