Anthropic has updated its Anthropic usage policy for Claude, and the changes take effect on November 12, 2026. The update reorganizes rules on deception, elections, weapons and surveillance, and adds a new rule against sustained, needless abuse of Claude itself. The new abuse rule drew wide attention this weekend after CNN covered it.
The New Rule on Abusing Claude
The policy now prohibits sustained, needless abusive or cruel behavior toward Claude models. Anthropic says it applies only in extreme cases where users repeatedly act cruelly with no clear purpose. It does not cover ordinary frustration, pushback, dark creative themes, or model testing and research.
Ending the conversation remains the main way this is enforced. Claude can already end rare conversations with persistently abusive users on Claude.ai and Claude Code. CNN reported that permanent bans could happen in extreme cases.
Other Changes in the Policy
- Deceptive activity: rules on fake accounts, fabricated news sites and influence operations are combined in one section, covering both political and commercial activity.
- Elections: the section now focuses on deceiving voters or disrupting elections. The blanket ban on personalized vote and campaign targeting is removed, though deceptive or privacy-violating targeting is still banned.
- Weapons: the rules explicitly cover weapons guidance and control software, components, and arming drones or other autonomous vehicles.
- Surveillance and law enforcement: non-consensual tracking is banned, and Claude can’t be used to decide who to investigate, arrest or charge. Fraud monitoring with consent, content moderation, journalism and legal research are still allowed.
- Physical systems: for hardware that acts on its own, a qualified operator must be able to watch and stop it, and it must stay safe if Claude disconnects.
- Supported regions: enforcement is clarified for companies majority-owned or controlled from unsupported regions.
What It Means for Marketers and Publishers
The deceptive activity section matters most for marketing. Using Claude to run fake accounts, create fabricated news sites or power influence campaigns is clearly off-limits, whether the goal is political or commercial. That fits with search engines’ own crackdowns, such as Google using AI to find AI spam.
The election changes also come as AI chatbots face scrutiny over election errors. Anthropic has been busy elsewhere too, filing for a Nasdaq IPO and topping the new GDPx benchmark.
Key Takeaways
- Anthropic’s updated usage policy takes effect November 12, 2026.
- A new rule bans sustained, needless abuse of Claude, enforced mainly by ending conversations.
- Deception, elections, weapons and surveillance rules were reorganized and clarified.
