Anthropic recently cut off OpenAI’s access to Claude’s API, accusing the company of utilizing Claude Code to refine its forthcoming GPT-5. The Amodei brothers emphasized that this action aligns with their terms of service, which forbid using the service to develop competing products or reverse-engineer it. This shift marks a significant change in AI ethics, where learning from competitors is now viewed as industrial espionage. The competitive landscape reveals an imbalance, with OpenAI valued at $300 billion following an $8.3 billion funding round and overseeing over 800 million weekly users. In contrast, Anthropic, valued between $61.5 and over $100 billion, positions Claude as a strong contender, boasting a 79% success rate in SWE-bench, outperforming GPT-4.1. As OpenAI prepares to launch GPT-5, the motivation behind its engineers’ interest in Claude’s capabilities is evident, illustrating the fierce rivalry in AI development.
Source link
