Anthropic has withdrawn OpenAI’s access to its Claude AI models, as reported by Wired. Sources revealed that OpenAI was integrating Claude with internal tools to evaluate its performance against OpenAI’s models in coding, writing, and safety metrics.
An Anthropic spokesperson told TechCrunch, “OpenAI’s technical team used our coding tools before GPT-5’s launch,” which breached their terms of service prohibiting the use of Claude to develop competing services. Nonetheless, Anthropic will still allow OpenAI access for benchmarking and safety evaluations.
In response, OpenAI labeled its use as “industry standard” and expressed disappointment, noting their API remains open to Anthropic. Anthropic previously resisted providing competitor access, with Chief Science Officer Jared Kaplan justifying cutting off Windsurf by saying, “It would be odd for us to sell Claude to OpenAI.”