Anthropic has reportedly prohibited OpenAI from accessing its Claude models.
Anthropic banned OpenAI after saying the developer had violated its terms of service, WIRED reported.
OpenAI employees were reportedly using the Claude models to measure the performance of its own models. Using certain application programming interfaces (APIs), OpenAI was allegedly integrating Claude into its own technology. This allowed the company to see how its models compared to Claude in areas such as creative writing and coding.
This testing would help OpenAI improve their models behaviour and how they respond to prompts.
Anthropic’s terms of service state that its technology cannot be used to build or improve other AI products, and yet the company itself is facing legal challenges about its own use of data. Just last month, Reddit accused Anthropic of scraping their platform’s content.
“Claude Code has become the go-to choice for coders everywhere, and so it was no surprise to learn that OpenAI’s own technical staff were also using our coding tools ahead of the launch of GPT-5.” said Christopher Nulty, Anthropic spokesperson, to WIRED. “Unfortunately, this is a direct violation of our terms of service.”
Claude Code, released in February 2025, explains codebases using agentic search.
OpenAI’s GPT-5 model is reportedly being released soon, with increased promotion from chief executive officer Sam Altman. The model is projected to surpass its predecessor, GPT-4, in intelligence and efficiency.
