Anthropic Updates Claude Usage Policy to Prohibit Unauthorized Model Distillation and Malicious Cyberattacks

Anthropic announced an updated Usage Policy, effective November 12, 2026, covering Claude, Claude Code, its API, and third-party integrations. The policy bans unauthorized use of model outputs to train other AI models, including scraping and distillation, and prohibits using Claude for malicious cyberattacks, malware development, and fraud. For high-risk applications in finance, healthcare, legal, and hiring, Anthropic requires qualified personnel to review AI recommendations and disclosure of AI involvement to affected parties. Consumer-facing chatbots and AI agents must also clearly disclose their AI identity. The move signals tightening control over how Claude's outputs are commercially reused, with implications for competitors relying on distilled models. Watch for enforcement details and how rival AI firms respond to the distillation restrictions.