Anthropic’s repeated disclosures of AI misuse may be doing more than informing the public: they may be forcing frontier AI companies toward a new competitive norm in which credibility depends on showing what went wrong, what was stopped, and what remains unknown.