Anthropic updated Claude’s rules. Starting November 12, 2026, prolonged abusive treatment of the models may result in access being restricted or terminated. This is stated in the company’s official blog, UNN reports.
Anthropic has updated the rules for using its artificial intelligence models. The document will take effect on November 12, 2026. The company prohibits prolonged and unfounded abusive or cruel treatment of neural networks
If a user is suspected of violating the rules, they may receive a warning. The company also reserves the right to restrict, suspend, or completely terminate access. The document does not establish automatic blocking for every individual insult.
The rules cover users of Claude.ai and Claude Code, developers and businesses working through the API, as well as customers of third-party products with Claude integrated.
In addition to conduct toward the models, the document regulates other ways of using the services. In particular, the following are prohibited:
- developing weapons and software for their use;
- organizing influence campaigns through fake accounts;
- monitoring people without their consent;
- circumventing blocks by using other accounts.
Anthropic reported a user to the police over a chat with Claude06.10.26, 10:28