Anthropic adds a rule against being cruel to Claude

Anthropic, the company that makes the AI called Claude, has changed its rules for using Claude. The company announced the change on October 8, 2026. The new rules start on November 12, 2026.

Anthropic updates these rules every year. This year, it added a new rule. People may not keep being cruel to the AI again and again when there is no real reason for it.
The rule is only for extreme cases. It does not apply to people who complain about the AI or disagree with it. It does not apply to dark themes in stories. It also does not apply to people who test or study the AI.
In Claude.ai and Claude Code, Claude can already end a conversation with a person who keeps being cruel. This will stay the main way the rule is used.
The update also made other rules clearer. These rules are about illegal activities, elections, weapons, watching people and police work, very risky uses, and the countries where people can use Claude.
Glossary
- 169 words
- 14 sentences
- 2 mins to read




