Anthropic Updates Its Usage Policy on Cruel Behavior Toward Models, Effective November 12

Anthropic published an updated Usage Policy on 8 October 2026. The updated policy takes effect on 12 November. As of 10 November 2026, it has not yet taken effect.

What the news post says

Anthropic's official news post describes a prohibition on sustained and needless abusive or cruel behavior toward its models, and says this applies only in extreme cases, where users repeatedly act cruelly toward the models, with no discernible purpose.

The same news post explains that this does not apply to common versions of user frustration, pushback, dark creative themes, or model testing and research. That exclusion is official news commentary; it is not the wording of the Usage Policy itself.

The news post also says Anthropic has already taken a step: allowing Claude models to end rare conversations with persistently abusive users on Claude.ai and Claude Code, and that Claude's ability to end these interactions remains the primary enforcement mechanism. That description is limited to Claude.ai and Claude Code.

What the Usage Policy says

The Usage Policy text is shorter. Under “Do Not Engage in Cruel, Abusive, or Psychologically Harmful Conduct,” it prohibits this: “Engage in sustained and needless abusive or cruel behavior toward our models.” The policy text does not include the news post's “extreme cases” or “no discernible purpose” language.

A general rule in the Usage Policy, not limited to this conduct section, says that encountering a block does not by itself indicate a violation of the Usage Policy.

If Anthropic suspects that a user may have violated the Usage Policy, it may warn the user or throttle, limit, suspend, or terminate access.

The policy applies to anyone who submits inputs to Anthropic's products or services. That includes individuals using its apps (such as Claude.ai and Claude Code), developers and businesses using the API and developer platforms, customers accessing Claude through cloud providers and authorized resellers, and end users of products or services that integrate Claude.

The United States version of the Usage Policy already shows this text under an effective date of 12 November 2026. A downloadable policy PDF carries the same effective date and the same prohibition wording.

A broader policy refresh

The 8 October news post is not only about cruel behavior toward models. It also updates sections on deceptive activities, elections and democratic processes, weapons (including guidance and control software and armed drones), surveillance and law enforcement, human-in-the-loop review in high-risk settings, physical devices related to the Model Hardware Standard, and Supported Regions. The cruel-behavior rule is one part of a larger policy refresh.