GoKawiilAnthropic updates Claude usage policy to ban sustained cruelty toward the AI
Anthropic has revised its usage policy for the first time in over a year, adding a rule against 'sustained and needless abusive or cruel behavior' toward its Claude models. The company clarified the rule targets only extreme, repeated mistreatment, not normal frustration, pushback or dark creative writing, and enforcement can range from warnings to account suspension or termination. The update also tightens rules on fake-account propaganda networks, weapons software, surveillance, and police-related recommendations, while removing a previous blanket ban on personalized voter and campaign targeting.
GoKawiil's interpretation of the reporting above, not reported fact.
The move builds on a feature Anthropic launched in August 2025 letting Claude end abusive conversations, which the company has framed as part of its research into AI welfare rather than pure safety policy. Formalizing this into usage rules suggests Anthropic is trying to establish enforceable norms around how humans treat AI systems, a question other AI labs haven't yet addressed as explicitly. The loosened stance on political targeting could also signal a shift in how Anthropic balances restrictions against commercial or political use cases ahead of future elections.
- Anthropic now bans sustained, needless cruelty toward Claude, enforceable via warnings up to account termination.
- The rule extends an August 2025 feature allowing Claude to end abusive conversations, tied to AI welfare research.
- Anthropic also clarified bans on fake-account propaganda and police-arrest recommendations while dropping its ban on personalized political targeting.
Source: techspot.com — Rob Thubron, 2026-10-09
Published there as: “Being mean to AI could get you banned as Anthropic introduces new rules against abusing Claude”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.