Anthropic has announced that it has revoked OpenAI's access to its Claude series of AI models. This news was first reported by Wired, with sources revealing that OpenAI had previously used an internal tool to connect to Claude in order to compare its performance in areas such as programming, writing, and security.
Image source note: The image is AI-generated, and the image licensing service is Midjourney
A spokesperson for Anthropic stated that OpenAI's technical team had used their coding tools before the release of GPT-5, which was considered a "direct violation of our terms of service." Anthropic's commercial terms clearly prohibit other companies from using Claude to build competitive services. Nevertheless, Anthropic said it would still provide some access to OpenAI for benchmark testing and security assessments.
In contrast, OpenAI responded in a statement that its use of the service was in line with industry standards and expressed disappointment at Anthropic's decision to cut off API access. At the same time, OpenAI pointed out that Anthropic's API remains open to them, showing a delicate relationship between the two companies.
Facts show that senior executives at Anthropic have previously shown resistance to granting access to competitors. Jared Kaplan, Anthropic's Chief Science Officer, had previously stated that selling Claude to OpenAI was unreasonable, a stance that attracted widespread attention in the industry.
Key points:
🌟 Anthropic has canceled OpenAI's access to the Claude model, citing a violation of the terms of service.
⚖️ OpenAI claims its use of Claude complies with industry standards and expresses disappointment.
🚀 The incident has intensified competition in the artificial intelligence industry and attracted widespread attention.