Anthropic bans needless abuse towards Claude in usage policy

Anthropic has recently updated its usage policy, prohibiting users from subjecting Claude, its artificial intelligence model, to sustained and needless abusive or cruel behavior. This new rule will apply only in extreme cases where users repeatedly act cruelly without a discernible purpose. The company clarifies that this does not include ordinary user frustration, pushback, dark creative themes, or model testing.

Why it matters

This update is significant because it reflects a growing concern for ethics in the use of artificial intelligence. By establishing clear limits on acceptable behavior towards Claude, Anthropic aims not only to protect its model but also to encourage more constructive and respectful interactions in the realm of AI. This could have broader implications for how human-machine interactions are developed and regulated in the future.

What we know

Claude is already capable of ending conversations with users who exhibit persistently abusive behavior. When this occurs, the user will not be able to send further messages in that conversation, although other chats will remain unaffected. This conversation-ending tool was initially introduced in August 2025 as part of research on AI welfare.

What remains unclear

There is still uncertainty regarding Claude's moral status, a topic that Anthropic is exploring. Although the company has indicated that it is unclear whether Claude possesses any moral standing, the implementation of these policies suggests a proactive approach to ethics in artificial intelligence.