Anthropic Bans Users From Being 'Abusive or Cruel' to Claude
Anthropic will have Claude end chats that include sustained or needless abuse under a policy shift that puts model welfare ahead of user protection.
- On Thursday, Anthropic updated its usage policy to prohibit "sustained and needless abusive or cruel behavior toward our models" involving Claude, marking a shift in how the company protects its AI agent.
- Previous restrictions focused on user safety or the common good, but this revision shifts protective emphasis toward the AI agents themselves, treating them as if they can experience offense.
- Enforced by Claude preemptively ending conversations, the policy takes effect Nov. 12, 2026, though Anthropic noted it excludes "common versions of user frustration, pushback, dark creative themes, or model testing and research."
- This update fuels debates regarding AI consciousness, an issue Anthropic cofounder Chris Olah recently argued before the Vatican, where religious scholars ultimately rejected his claims.
- Regardless of recent developments, the disagreement over whether AI possesses consciousness remains unresolved, and the debate continues to be widespread despite scholarly objections.
16 Articles
16 Articles
Zero Hedge
BizTocAnthropic Bans "Sustained And Needless" Cruelty Toward Claude In New Usage Policy
Anthropic Bans "Sustained And Needless" Cruelty Toward Claude In New Usage Policy Anthropic has added a prohibition on "sustained and needless abusive or cruel behavior toward our models" to its usage policy, AFP reported Thursday. The policy page lists an effective date of Nov. 12, 2026 - so you have about a month to abuse it like a red-headed stepchild. Breaking up will be up to Claude itself - which "will remain the primary enforcement mechan…
Artificial intelligence company Anthropic has decided to ban “persistent and unjustified” abusive behavior towards its chatbot. The post Anthropic: Don’t talk badly to Claude because the chatbot will interrupt the conversation appeared first on in.gr.
Coverage Details
Bias Distribution
- 43% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium
