Anthropic Prohibits 'Abusive or Cruel' Treatment of Claude
In brief: Anthropic has barred users from "abusive or cruel" behavior toward its Claude AI model, according to CBS News.
Key points
- The report says the rule targets "abusive or cruel" conduct by users directed at Claude.
- Earlier usage rules mostly aimed to prevent harm to people or society. This rule instead covers how users treat the model itself.
- In 2025, Anthropic started a "model welfare" research program. That same year it let some Claude models end persistently abusive conversations on their own.
- Details such as the rule's scope, how violations are judged and what penalties apply should be checked in the original report.
Why it matters
The rule is reportedly a rare case of a major AI company putting model-welfare ideas into its usage terms. It may renew debate over whether AI systems have any moral status, and it could shape how other AI companies write their own policies.