OpenAI has had its access to Anthropic’s ‘Claude’ API revoked, after it was found to be violating the terms of service.
Anthropic’s terms prohibit the use of the service to build rival products or services, something that OpenAI has been accused of doing in relation to its soon-to-be-launched ‘GPT-5’ model.
GPT-5 is rumoured to be more efficient at coding, which may explain OpenAI’s need to test Claude’s capabilities in advance of the model’s launch.
Anthropic’s decision to cut OpenAI’s access, despite citing it as standard practice across the industry for benchmarking and safety evaluation, may be a result of the company acquiring AI coding startup Windsurf so that it could access its tech.
The deal fell through after tensions between the two companies arose.
