Anthropic's updated usage policy bans sustained and needless cruelty toward its Claude models, with Claude ending abusive chats as the main enforcement
Anthropic Bans "Sustained And Needless" Cruelty Toward Claude In New Usage Policy
Context
The rule is a rare case of an AI company writing protections for the model itself into its terms, which feeds the debate over whether AI systems could have interests worth protecting.
ZeroHedge is relaying a story AFP first reported on October 8, so this is a few days old; wongmjane also flagged the change on X on October 8. Anthropic announced the change in its 2026 usage policy update. ZeroHedge reported that the policy page lists an effective date of November 12, 2026.
Deep dive
On October 12, zerohedge posted: Anthropic Bans "Sustained And Needless" Cruelty Toward Claude In New Usage Policy, linking its own article.
Anthropic's update says it has added a prohibition on sustained and needless abusive or cruel behavior toward its models. It says the rule applies only in extreme cases, where users repeatedly act cruelly toward the models with no discernible purpose, and not to ordinary frustration, pushback, dark creative themes, or model testing and research. It ties the change to an earlier step letting Claude end rare conversations with persistently abusive users on Claude.ai and Claude Code, and says that ability will remain the primary enforcement mechanism.
AFP reported that the policy does not explicitly cite model welfare, the idea that AI systems might deserve protections usually reserved for living things, but that Anthropic and its executives have openly entertained it. ZeroHedge quoted CEO Dario Amodei telling The New York Times in February that the company does not know if the models are conscious but is open to the idea, and Microsoft AI chief Mustafa Suleyman writing last month that AIs are not conscious.
The rest of the ZeroHedge piece is commentary, including an essay arguing that talk of machine consciousness serves AI companies commercially; it notes that nothing in AFP's account supports that reading.
What to expect next
Watch how Anthropic applies the rule after the November 12 effective date, including whether any accounts are restricted for abuse rather than conversations simply being ended.
Also watch whether OpenAI, Google or other labs add similar language, and whether the change fuels more debate over model welfare from critics like Mustafa Suleyman and from religious leaders who have rejected machine consciousness.