2026 Usage Policy update

What it claims

Anthropic's Usage Policy update takes effect on November 12, 2026, alongside a revised Supported Regions Policy. Anthropic says most changes clarify existing rules rather than change enforcement. The weapons and surveillance rewrites are stated to match how the previous policy was already enforced. The genuinely new obligations are hardware controls, a ban on abusive behavior toward models, and the removal of the blanket ban on personalized vote and campaign targeting.

The new section, "Do Not Engage in Deceptive Campaigns or Artificial Activity," applies to political and commercial activity alike. It covers hiding who is behind a message, amplifying content through fake accounts or posts, and building tools or infrastructure for influence operations. The elections section is renamed "Do Not Undermine Democratic Processes." It now targets false information about candidates or voting, impersonating candidates or officials, and suppressing turnout. Anthropic cites nonprofits writing voter information in other languages and election officials sending ballot cure notices as legitimate work the old targeting ban caught. Voter targeting that uses deception or misuses personal data stays prohibited under the deceptive campaigns, surveillance and privacy sections.

How to use it

Teams building on Claude should check their product against these requirements.

• Weapons: The ban covers guidance and control software, the components that make weapons work, and arming drones or other autonomous vehicles.
• Surveillance and criminal justice: Tracking people without consent is banned, in real time or through analysis of previously collected data. So is building or improving surveillance tools. Claude cannot decide or recommend who to investigate, arrest or charge. The policy explicitly permits consented tracking such as fraud monitoring, content moderation, journalism and legal research.
• High-risk uses (health, legal rights, finances, livelihood, access to essential services): A qualified human in the loop must have authority to review and change Claude's recommendations. The affected person must be told that AI was used. These rules are unchanged, but the rewritten section now lists which kinds of recommendations are covered and which are not.
• Physical hardware: This applies when Claude drives hardware that takes autonomous physical actions and could cause injury. A qualified operator must be able to observe the equipment and stop it. The equipment must hold a safe state if Claude disconnects. The rule follows Anthropic's Model Hardware Standard.
• Supported regions: Use is barred for people physically located in an unsupported region and for entities incorporated or headquartered in one. It is also barred for entities majority-owned or controlled by persons or entities in those regions.

Limits

The abuse prohibition targets only "sustained and needless" cruelty toward models with no discernible purpose. It excludes ordinary frustration, pushback, dark creative themes, and model testing or research. The main enforcement mechanism remains Claude's ability to end conversations with persistently abusive users on Claude.ai and Claude Code.

The post summarizes changes and does not reproduce the policy text. The exact list of covered high-risk recommendations, the list of unsupported regions, and the Model Hardware Standard sit on separate pages. Teams should read those pages before relying on any specific interpretation.