Anthropic has revoked OpenAI’s access to its Claude family of AI models, according to a report in Wired.

Sources told Wired that OpenAI was connecting Claude to internal tools that allowed the company to compare Claude’s performance to its own models in categories like coding, writing, and safety.

TechCrunch has reached out to Anthropic and OpenAI for comment. In a statement to Wired, an Anthropic spokesperson said, “OpenAI’s own technical staff were also using our coding tools ahead of the launch of GPT-5,” which is apparently “a direct violation of our terms of service.” (Anthropic’s commercial terms forbid companies from using Claude to build competing services.)

However, Anthropic also said it would continue to give OpenAI access for “benchmarking and safety evaluations.”

Meanwhile, an OpenAI spokesperson said, “While we respect Anthropic’s decision to cut off our API access, it’s disappointing considering our API remains available to them.”

Anthropic executives had already shown resistance to providing access to competitors, with Chief Science Officer Jared Kaplan previously justifying the company’s decision to cut off Windsurf (a rumored OpenAI acquisition target, subsequently acquired by Cognition) by saying, “I think it would be odd for us to be selling Claude to OpenAI.”