The quoted phrase is real and comes from Anthropic's own announcement. The official post is dated Oct 8, 2026, and opens by saying that each year Anthropic updates its Usage Policy in response to evolving model capabilities and customer feedback, and that a new version of the policy is being published that day . The post contains a section headed "Addressing abusive behavior toward our models," which states that the addition aligns with a step already taken, allowing Claude models to end rare conversations with persistently abusive users on Claude.ai and Claude Code, that such abuse is the main focus of this update, and that "Claude's ability to end these interactions will remain the primary enforcement mechanism." The same post states the carve-outs: it does not apply to common versions of user frustration, pushback, dark creative themes, or model testing and research .

On narrowness, Anthropic's own words as reported by the originating outlet: "The policy update is meant to apply only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose."

On timing, the prohibition is not yet operative. The rules take effect November 12, giving users just over a month . Independent reads of the live policy page agree: the new text is the version now served at Anthropic's usage policy page, which gave an effective date of November 12, 2026 , and the full text of the new policy is already live, marked "Effective November 12, 2026," with the previous version kept in an archive .

On where the clause sits in the document, one outlet reading the policy text reports that it lists "sustained and needless abusive or cruel behavior toward our models" under a section titled "Do Not Engage in Cruel, Abusive, or Psychologically Harmful Conduct," alongside prohibitions on harassing or bullying people and on promoting self-harm or graphic violence .

The conversation-ending feature predates this update. In August 2025 Anthropic announced capabilities allowing some of its largest models to end conversations in "rare, extreme cases of persistently harmful or abusive user interactions," and said it was doing this not to protect the human user but rather the AI model . Importantly for the consciousness framing: the company was not claiming its Claude models are sentient or can be harmed, and said it remains "highly uncertain about the potential moral status of Claude and other LLMs, now or in the future."

The cruelty clause is one of several changes. The same update consolidates scattered rules into a new section on deceptive activity after Anthropic observed state media outlets, government propaganda offices and commercial firms using Claude to run networks of fake accounts and fabricated news sites, and rewrites the surveillance and criminal justice section to state more precisely what is prohibited, including a prohibition on using Claude to build or improve tools designed for surveillance .

On the two attributed consciousness positions in the post's caption: Anthropic's CEO told The New York Times in February that he was unsure whether AI models could be conscious, saying "We don't know if the models are conscious...But we're open to the idea that it could be." On the Microsoft side: Microsoft AI CEO Mustafa Suleyman told CNBC in late 2025 that he believes consciousness can only occur in biological beings, and he has become one of the most prominent industry voices against building seemingly conscious AI . In his own earlier essay he framed the issue this way: he said he had become increasingly concerned about people believing so strongly in AIs as conscious entities that they would advocate for AI rights, calling this a dangerous turn that must be avoided .