Anthropic has cut off OpenAI’s access to its Claude AI models, Wired reports.
The problem: OpenAI connected Claude to its internal tools. It used Claude to compare performance against its own models in coding, writing, and safety. This broke Anthropic’s terms of service.
Anthropic’s spokesperson told Wired: 
“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 terms forbid using Claude to build competing services. However, Anthropic says some access will continue for “benchmarking and safety evaluations.”
OpenAI pushed back, calling its usage “industry standard” and added:
“While we respect Anthropic’s decision to cut off our API access, it’s disappointing considering our API remains available to them.”
Anthropic has been wary of sharing Claude with competitors. Chief Science Officer Jared Kaplan previously said about cutting off Windsurf (a rumored OpenAI acquisition target):
“I think it would be odd for us to be selling Claude to OpenAI.”
