Anthropic bans users from being 'cruel' to AI — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic updated its usage policy to ban users from "sustained and needless" abusive behavior toward its AI, letting its tools end "cruel" interactions — something the company said it had previously done only in "rare" cases
- The new "cruel" provision was added to a list of forbidden conduct that already covered bullying others, promoting self-harm, and creating non-consensual intimate imagery
- Anthropic said the policy excludes common user frustration, pushback, dark creative themes, and model testing — applying only to "extreme cases" of repeated abuse
- The update, first reported by The Verge, also added language barring "deceptive campaigns" and use of Claude to "deceive voters or disrupt elections" ahead of US midterms, and reaffirmed the company's existing ban on developing weapons
- Microsoft AI director Mustafa Suleyman, who criticised Anthropic in September for treating AI as if it were human, reacted to the announcement with a melting-face emoji
- Irish tech lawyer Barry Scannell called the move "deeply wrong," posting on LinkedIn: "You can't be cruel to numbers and maths. This level of anthropomorphisation of AI is harmful"
- OpenAI's Sam Altman previously framed user politeness — the "please" and "thank you" messages — as "tens of millions of dollars well spent," a thread the article ties to the broader debate over how to speak to AI
Why it matters: By listing cruelty to a chatbot alongside bullying and self-harm as forbidden conduct, Anthropic formally elevated AI into a category of thing that can be harmed — a framing that prompts a direct rebuke from Microsoft's AI chief and a legal-academic warning that the anthropomorphization "leads people to believe it's something it's not." Whatever the policy's actual enforcement, the public posture splits the industry's biggest players at a moment when etiquette toward generative AI is becoming a consumer-level question.
Ask SkimNews
