Anthropic Revokes OpenAI's API Access for Terms-of-Service Violations
Anthropic has revoked OpenAI's access to the Claude API, with the suspected cause being unauthorized distillation of Claude's outputs—a violation of Anthropic's terms of service. The story was surfaced by Hugging Face CEO Clément Delangue and reported by Wired. The revocation comes amid a broader industry tension over training data practices: Delangue has separately argued that major labs trained on web data distillation and are now deploying legal resources to prevent competitors from doing the same.
Why It Matters
A direct API access revocation between two frontier labs is unprecedented and signals that the training data and model distillation conflict—previously fought in courts and public opinion—has crossed into operational infrastructure enforcement.