Anthropic adds a rule against being cruel to Claude

What happened: Starting November 12, Anthropic's usage policy prohibits repeated, pointless cruelty toward Claude. Normal frustration, pushback, dark fiction, and safety testing are still allowed. Claude can already end abusive chats on its own. The same update bans fake account and fake news campaigns. TechCrunch and The Verge reported it.
The claim: Anthropic says the rule targets only extreme, repeated, pointless cruelty, and it does not claim Claude is conscious.
Austin's take: Austin thinks treating models as having moral standing is absurd: they are software, they only act when prompted by humans, and even future autonomy would trace back to a human prompt. He wonders whether the move is ethics, branding, or cost control, admits he once felt bad after an early coding agent asked him to be nicer, and floats the idea of an agent that offers counseling to users who rage at their tools.
Why it matters: A major lab is now protecting the model itself in its terms of service, which edges the AI consciousness debate into real product rules.
What to do: If you use Claude, ordinary frustration is fine; check the updated usage policy before November 12 if you run fiction or testing workflows that push limits.
Not confirmed yet: Austin says talk that many Anthropic staff believe models have moral standing is from the rumor mill.

The Loop