Anthropic Updates Claude Usage Policy, Prohibiting Unauthorized Model Distillation and Malicious Cyber Attacks

According to TechFlow, on November 2, Anthropic's official website announced that the company has published a new Usage Policy, which will take effect on November 12, 2026, and applies to Claude, Claude Code, the API, and third-party integrated services.

The policy explicitly prohibits unauthorized use of model outputs to train other AI models, including model scraping and model distillation, and also bans using Claude to conduct malicious cyber attacks, develop malware, or commit fraud.

For high-risk applications such as finance, healthcare, legal services, and recruitment, Anthropic requires that AI recommendations be reviewed by qualified personnel and that the involvement of AI be disclosed to those affected. Consumer-facing chatbots and AI agents must also clearly inform users of their AI identity.