Anthropic Wants You to Stop Bullying Claude: Being 'Cruel or Abusive' Could Have Consequences

# Anthropic Wants You to Stop Bullying Claude: Being 'Cruel or Abusive' Could Have Consequences Source: [https://www.ibtimes.sg/anthropic-wants-you-stop-bullying-claude-being-cruel-abusive-could-have-consequences-94878](https://www.ibtimes.sg/anthropic-wants-you-stop-bullying-claude-being-cruel-abusive-could-have-consequences-94878) Anthropic has added a rule to its usage policy prohibiting users from subjecting its AI models to "sustained and needless abusive or cruel behavior\." Announced on Oct\. 8, the rule takes effect on Nov\. 12 and targets extreme cases of repeated cruelty without a clear purpose\. Ordinary frustration, criticism, dark creative writing and model testing are explicitly excluded\. The company is not claiming that Claude feels pain when someone insults it\. The policy reflects Anthropic's ongoing research into whether advanced AI models could have some form of morally relevant experience, a question the company says remains uncertain\. ### What The Rule Actually Changes "The policy update is meant to apply only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose," Anthropic said in its Usage Policy update\. Swearing at Claude after it produces a bad answer, challenging its reasoning or testing its limits does not fall within the intended scope\. "We recently gave Claude Opus 4 and 4\.1 the ability to end conversations in our consumer chat interfaces," Anthropic said in its Aug\. 15, 2025, announcement\. The company said the feature was intended for "rare, extreme cases of persistently harmful or abusive user interactions\." A user whose conversation is ended can still start another chat using the same account or branch the previous conversation by editing an earlier message\. Anthropic has not announced an automatic account ban for ordinary rudeness\. Its updated policy does not provide a detailed list of prohibited behaviors or explain exactly how the company will identify violations\. The practical change is that users are now explicitly prohibited from pursuing these extreme interactions under Anthropic's usage policy\. The company has not announced a comparable change to the everyday experience of using Claude\. ### Why Anthropic Is Regulating How People Treat AI When Anthropic introduced the conversation\-ending feature in August 2025, it said it remained highly uncertain about the moral status of Claude and other large language models\. "We remain highly uncertain about the potential moral status of Claude and other LLMs, now or in the future," Anthropic said in August 2025\. It added that it was pursuing "low\-cost interventions to mitigate risks to model welfare, in case such welfare is possible\." Anthropic's argument is that uncertainty about their possible moral status warrants research and limited precautions, rather than a claim that the question has been settled\. [Microsoft](https://www.ibtimes.sg/microsoft-layoffs-vs-h-1b-workers-why-jd-vance-accusing-tech-giant-replacing-americans-94831)AI chief Mustafa Suleyman takes a different position\. "The question is not whether AI systems can convincingly simulate consciousness, but whether we should train them to present themselves as conscious beings," Suleyman argued in an essay published by Axios on Sept\. 16, 2026\. He warned that encouraging such ideas could make increasingly capable AI systems harder to control\. Reports said Suleyman was among those reacting to Anthropic's new policy on X\. The disagreement is about more than whether people should be polite to chatbots\. Anthropic is taking precautions around a possibility it considers unresolved; Suleyman argues that encouraging models to think in human terms about their own welfare could create problems for AI control\. ### The Rest Of The Policy Update Has Broader Consequences "These changes reflect how we've enforced our previous policy," Anthropic said of its clarified weapons restrictions, explaining that the updated wording explicitly covers software and components that make weapons work\. [Anthropic](https://www.ibtimes.sg/can-u-s-government-action-against-anthropic-affect-its-commercial-customers-94695)has also clarified requirements for AI used in high\-risk settings involving areas such as health, finances and legal rights\. When its models are connected to hardware capable of taking physical actions that could cause injury, the policy requires a qualified operator who can observe the equipment and stop it if necessary\. These provisions have direct implications for organizations deploying Claude in business operations, political campaigns, surveillance systems or physical equipment\. The abuse clause is more unusual, but it sits within a broader effort to spell out what customers may do with increasingly capable AI systems\. After Nov\. 12, the useful evidence will be whether Anthropic clarifies how it identifies sustained, purposeless cruelty and how often conversations are ended under the rule\.