Anthropic has revoked OpenAI’s access to its Claude AI models due to a violation of terms of service that prohibit using the service to create competing products. OpenAI was reportedly using Claude for benchmarking against its own models, including the upcoming GPT-5 release. Anthropic spokesperson Christopher Nulty confirmed the situation, highlighting that Claude Code is popular among coders. Although OpenAI’s actions were justified as “industry standard,” the loss of API access is disappointing, especially since OpenAI’s API remains available to Anthropic. This incident mirrors past actions by tech companies, showcasing a pattern of restricting competitor access to protect intellectual property. OpenAI’s CEO Sam Altman expressed concerns about the highly advanced GPT-5, likening its testing to the development of the Manhattan Project. This development is significant for AI performance and safety evaluations, positioning both companies in crucial market dynamics as the AI landscape evolves.
Source link
