Anthropic has revoked OpenAI’s access to its Claude family of artificial intelligence models, according to a report in Wired.
Sources told Wired that OpenAI connected Claude to internal tools that allowed the company to compare Claude’s performance to its own models in categories such as coding, writing, and security.
In a statement provided to TechCrunch, an Anthropic spokesperson said: “OpenAI’s own technical staff also used our coding tools prior to the launch of GPT-5,” which is apparently “a direct violation of our terms of service.” (Anthropic’s terms of service prohibit companies from using Claude to create competing services.)
However, the company also stated that it will continue to provide access to OpenAI “for benchmarking and security assessment purposes.”
Meanwhile, in its own statement, an OpenAI spokesperson called its use an “industry standard” and added: “While we respect Anthropic’s decision to stop accessing our API, it is disappointing given that our API remains available to them.”
Anthropic executives have previously demonstrated resistance to providing access to competitors, and Chief Scientist Jared Kaplan previously justified the company’s decision to disable Windsurf (rumored to be a target of OpenAI, which was later acquired by Cognition) by saying, “I think it would be weird if we were selling Claude to OpenAI.”
