Anthropic revoked OpenAI,s access to its Claude API, accusing it of violating terms of service by using Claude's tools to help develop GPT-5. Anthropic claims OpenAI bypassed standard interfaces to run large-scale internal testing, including safety evaluations. While OpenAI acknowledges the activity, it defends it as standard industry practice for benchmarking.
This clash reveals deeper competitive tensions, following Anthropic's earlier block of Claude access to Windsurf, a startup OpenAI aimed to acquire. The feud underscores rising aggression in the AI arms race, with companies using API access as strategic leverage to limit rivals' advancements.
