Anthropic Revokes OpenAI’s Access to Claude AI Models
In a recent development reported by Wired, Anthropic has decided to revoke OpenAI’s access to its Claude family of AI models. This decision comes after sources revealed that OpenAI was utilizing Claude in a manner that violated Anthropic’s terms of service.
According to the sources, OpenAI had been connecting Claude to internal tools to compare its performance with their own models in various areas such as coding, writing, and safety. This unauthorized usage led to Anthropic taking action to restrict OpenAI’s access to Claude.
An Anthropic spokesperson stated to JS that OpenAI’s technical staff were using their coding tools prior to the launch of GPT-5, which infringed upon the terms of service prohibiting the use of Claude to develop competing services. Despite this, Anthropic has decided to allow OpenAI to continue accessing Claude for benchmarking and safety evaluations.
On the other hand, an OpenAI spokesperson defended their actions by describing their usage as “industry standard” and expressing disappointment over the decision to cut off API access. They emphasized that their API remains available to Anthropic, highlighting the ongoing partnership between the two companies.
This incident reflects Anthropic’s stance on providing access to competitors, with Chief Science Officer Jared Kaplan previously expressing reluctance to sell Claude to OpenAI. The company’s decision to restrict access to Claude underscores the importance of upholding terms of service and protecting intellectual property rights.
