Anthropic bans ‘abusive or cruel behavior’ toward Claude

Anthropic has revised its usage policy for Claude, the first update in more than a year, to address what the company describes as new and high-risk cases of misuse. The changes cover election interference, weapons development, surveillance, and health and financial uses, according to the company.

One of the most significant additions prohibits "sustained and needless abusive or cruel behavior" toward Claude. The update follows an announcement last August, when Anthropic said it would let Claude end conversations with users it described as "persistently harmful or abusive," framing that step as part of its research into "model welfare."

Under the new policy, terminating conversations remains the "primary enforcement mechanism" for such behavior, the company said. Anthropic did not provide a comment on the update.

Why it matters: the policy sets expectations for how developers and product teams building on Claude can expect enforcement to work, with conversation termination rather than another mechanism serving as the main response to abuse directed at the model.