Anthropic’s 2026 Usage Policy update takes effect on November 12, sharpening the rules on deceptive campaigns, elections, weapons, surveillance, high-risk use cases, abusive behavior toward Claude, and supported-region access. Most of the changes clarify existing rules rather than introduce new prohibitions, and they reflect how Claude is now being used for longer, more independent work.<\/p>\n
Anthropic publishes a Usage Policy refresh each year to keep pace with model capability changes and customer feedback. In the year since the last refresh, Claude has taken on longer, more autonomous tasks. The 2026 version adds examples that show how the existing rules apply to those new capabilities and draws on new patterns of misuse observed over the past year, including influence operations, weapons development, and surveillance, as documented in Anthropic’s latest threat intelligence report.<\/p>\n
Past policy versions scattered rules against deceptive activity across the elections, fraud, privacy, and disinformation sections. The 2026 update consolidates them into a single section titled “Do Not Engage in Deceptive Campaigns or Artificial Activity,” which applies to deceptive activity of any kind, whether political or commercial.<\/p>\n
The section covers efforts to obscure who is behind a message, amplifying content through fake accounts or posts, and building the tools and infrastructure for running influence operation campaigns. Anthropic’s September threat intelligence report documented state media outlets, government propaganda offices, and commercial firms using Claude to run networks of fake accounts and fabricated news sites, which prompted the consolidation.<\/p>\n
The elections section has been renamed “Do Not Undermine Democratic Processes” to reflect its narrower scope. The renamed section specifically disallows using Claude to deceive voters or disrupt elections, including spreading false information about candidates or how to vote, impersonating candidates or election officials, or trying to suppress turnout.<\/p>\n
The update also removes the blanket prohibition on personalized vote and campaign targeting, which had covered legitimate civic work such as nonprofits writing voter information in other languages or election officials sending ballot cure notices. The behaviors that originally motivated the rule, including voter targeting built on deception or misuse of personal data, remain prohibited under the deceptive campaigns, surveillance, and privacy sections.<\/p>\n
The Usage Policy has always prohibited using Claude to develop weapons. Recent attempts to use the models to build guidance and control software for weapons prompted a clarification: prohibitions now explicitly include the software and components that make weapons work, as well as actions like arming drones and other autonomous vehicles. These changes match how Anthropic has enforced previous policy in practice.<\/p>\n
Recent threat intelligence work documented the growing use of AI to build surveillance systems that identify and track political dissidents. The rewritten surveillance and criminal justice section states more precisely what is prohibited.<\/p>\n
The section also lists which uses remain permitted, including tracking that people have agreed to (such as fraud monitoring), content moderation, journalism, and legal research.<\/p>\n
When Anthropic’s products are used in ways that can affect someone’s health, legal rights, finances, livelihood, or access to essential services, a qualified “human in the loop” must have the authority to review and, if necessary, change Claude’s recommendations, and the affected individual must be told that AI was used. These requirements are unchanged.<\/p>\n
Because users frequently asked whether a given use case met the requirements, the High-risk Use Case Requirements section has been rewritten to answer that directly. It now lists which kinds of recommendations are covered and which are not.<\/p>\n
Following the launch of Anthropic’s Model Hardware Standard, the section also adds requirements for use of the models when they are connected to hardware that takes autonomous physical actions and might be capable of causing injury. A qualified operator must be able to observe the equipment and stop it if needed, and the equipment must be able to hold a safe state if Claude is disconnected.<\/p>\n
A new prohibition covers sustained and needless abusive or cruel behavior toward Anthropic’s models. The rule applies only in extreme cases, where users repeatedly act cruelly with no discernible purpose, and does not apply to common user frustration, pushback, dark creative themes, or model testing and research.<\/p>\n
The policy update aligns with an already-active capability: Claude models can end rare conversations with persistently abusive users on Claude.ai and Claude Code. Ending these interactions remains the primary enforcement mechanism.<\/p>\n
The Supported Regions Policy was updated last year to restrict access to Claude for companies that are majority-owned by entities headquartered in unsupported regions. The Supported Regions page has now been clarified to spell out how the policy is enforced. It prohibits use of Claude by people physically located in an unsupported region, by entities incorporated or headquartered in one, and by entities majority-owned or controlled by persons or entities in those regions.<\/p>\n
The updated policy takes effect on November 12. The full Usage Policy and Supported Regions Policy are available on their respective pages. Anthropic has indicated it will continue updating the Usage Policy as Claude’s capabilities and risks change, seeking input from across the organization, policymakers, subject matter experts, civil society, and the people who use the products.<\/p>\n
The updated policy takes effect on November 12.<\/p>\n
The rules against deceptive activity, previously scattered across the elections, fraud, privacy, and disinformation sections, are now consolidated into a single section titled “Do Not Engage in Deceptive Campaigns or Artificial Activity,” applying to deceptive activity of any kind, whether political or commercial.<\/p>\n
The underlying prohibitions are unchanged. The update makes explicit that prohibitions cover the software and components that make weapons work, as well as actions like arming drones and other autonomous vehicles, and reflects how the previous policy was enforced in practice.<\/p>\n