AI news · Anthropic · policy
Anthropic bans sustained cruelty towards Claude
The new usage policy, in force from 12 November, also merges its rules on fake accounts into one section on deceptive campaigns.
Anthropic published a new version of its Usage Policy on Thursday 8 October 2026, and it adds a ban on “sustained and needless abusive or cruel behavior” towards its models, the company says. The policy takes effect on 12 November. Anthropic says the rule is for extreme, repeated cruelty with no purpose behind it; venting, arguing back, dark fiction and research on the models stay allowed.
Claude’s existing ability to end such conversations remains the main means of enforcement, Anthropic says. Anthropic gave no answer on whether bans could follow, The Verge reports, and the update is the first in over a year.
The rest of the policy changes more for business users. Rules on fake accounts and fabricated news sites, once spread across four sections, now sit in one section on deceptive campaigns that covers commercial as well as political activity, Anthropic says. The new rules also bar deceiving voters, TechCrunch notes. The elections section now targets voter deception and disruption, and a blanket ban on personalised vote and campaign targeting is gone. The weapons rules name guidance and control software and the arming of drones, and Claude connected to machines that act physically needs a qualified operator who can watch them and stop them.
Marketers using Claude should read the deceptive-campaigns section: it bars hiding a message’s real sender and boosting it with fake accounts or posts.
Sources
- Anthropic: 2026 Usage Policy update. Checked
- The Verge: “Anthropic bans ‘abusive or cruel behavior’ toward Claude”. Checked
- TechCrunch: Anthropic changes usage policy to ban model abuse and election interference. Checked
Alexandre S.
Alexandre S. is a creative ad strategist and copywriter with ten years' experience in digital marketing across multiple channels.