Anthropic Updates Usage Policy, Adding Model Abuse and Deception Rules

Anthropic published its annual 2026 Usage Policy update on October 8, 2026, consolidating its rules against deceptive campaigns, adding a prohibition on abusive behavior toward its models, and clarifying weapons, surveillance, and high-risk use requirements. The updated policy takes effect November 12, 2026.

Anthropic said most of the changes are intended to clarify existing rules. The company said the revision responds to a year in which Claude took on longer and more independent work, incorporates customer feedback, and draws on misuse patterns in influence operations, weapons development, and surveillance documented in its September 2026 threat intelligence report.

The Usage Policy, also called the Acceptable Use Policy, applies to anyone who can submit inputs to Anthropic’s products or services, including through authorized resellers or passthrough access. It is organized into Universal Usage Standards applying to all users, High-Risk Use Case Requirements for specific consumer-facing uses with elevated risk of harm, and Additional Use Case Guidelines covering consumer-facing chatbots, products serving minors, agentic use, and Model Context Protocol servers. Anthropic’s Safeguards Team implements detection and monitoring to enforce the policy; violations can bring throttling, suspension, or termination of access, and Anthropic may block or modify model outputs when inputs violate the policy.

Deceptive Campaigns and Elections Rules

Anthropic said it has seen state media outlets, government propaganda offices, and commercial firms use Claude to run networks of fake accounts and fabricated news sites. That activity was already prohibited, but the restrictions were scattered across the elections, fraud, privacy, and disinformation sections. The update consolidates them into a new section titled Do Not Engage in Deceptive Campaigns or Artificial Activity, which applies to deceptive activity of any kind, political or commercial. It covers efforts to obscure who is behind a message, amplification of content through fake accounts or posts, and building the tools and infrastructure for running influence operation campaigns.

The elections section is renamed Do Not Undermine Democratic Processes and focuses specifically on disallowing the use of Claude to deceive voters or disrupt elections, including spreading false information about candidates or how to vote, impersonating candidates or election officials, and suppressing turnout.

Anthropic also removed its blanket prohibition on personalized vote and campaign targeting. The company said the rule had covered legitimate civic work, such as nonprofits using Claude to write voter information in other languages and election officials sending ballot cure notices. Targeting that relies on deception or the misuse of voters’ personal data remains prohibited under the deceptive campaigns, surveillance, and privacy sections. The policy text currently in force, which remains in effect until November 12, 2026, still lists personalized vote and campaign targeting among prohibited political activities.

Weapons, Surveillance, and Law Enforcement

Anthropic said it has observed attempts to use its models to build guidance and control software for weapons. The updated weapons section makes explicit that the prohibitions include the software and components that make weapons work, as well as actions such as arming drones and other autonomous vehicles. Anthropic said the changes reflect how it has enforced the previous policy.

The rewritten surveillance and criminal justice section prohibits tracking people without their consent, whether in real time or through analysis of previously collected data. Claude cannot be used to decide or recommend who to investigate, arrest, or charge in a law enforcement or criminal justice process, or to build or improve tools designed for surveillance. Anthropic said its September report documented growing use of AI to build surveillance systems that identify and track political dissidents, and that the surveillance changes, like the weapons changes, make existing enforcement practice more explicit rather than changing it. The section also spells out uses that remain permitted, including tracking people have agreed to, such as fraud monitoring, content moderation, journalism, and legal research.

High-Risk Use Cases and Hardware Controls

When Anthropic’s products are used in ways that can affect someone’s health, legal rights, finances, livelihood, or access to essential services, the policy requires a qualified human in the loop — someone with authority to review and, if necessary, change Claude’s recommendations — and requires that the affected individual be told AI was used. Anthropic said these requirements have not changed, but it rewrote the section to answer a frequent user question more directly: it now lists which kinds of recommendations are covered and which are not.

Following the launch of Anthropic’s Model Hardware Standard research preview, the section adds requirements for models 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.

Abusive Behavior Toward Models

The update adds a prohibition on sustained and needless abusive or cruel behavior toward Anthropic’s models. The company said the rule is meant to apply only in extreme cases, where users repeatedly act cruelly toward the models with no discernible purpose, and does not apply to common user frustration, pushback, dark creative themes, or model testing and research.

The prohibition connects to a capability Anthropic introduced in 2025. In an August 15, 2025 research post, the company said it had given Claude Opus 4 and 4.1 the ability to end conversations in its consumer chat interfaces, intended for “rare, extreme cases of persistently harmful or abusive user interactions.” Anthropic said the feature was developed primarily as part of its exploratory work on potential AI welfare, with broader relevance to model alignment and safeguards.

According to that post, Claude is directed to use the ability only as a last resort, when multiple redirection attempts have failed or a user explicitly asks to end a chat, and not when a user might be at imminent risk of harming themselves or others. When a conversation ends, the user cannot send new messages in it, but other conversations on the account are unaffected, and previous messages can be edited and retried to create new branches. Anthropic said this ability, available on Claude.ai and Claude Code, will remain the primary enforcement mechanism for the new abuse prohibition.

Supported Regions Clarification

The update also addresses the Supported Regions Policy, which Anthropic tightened in 2025 for companies majority-owned by entities headquartered in unsupported regions. Anthropic said the clarified page explains how it enforces the policy: use of Claude is prohibited 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.

Anthropic said it will keep updating the Usage Policy as Claude’s capabilities and risks change, seeking input from policymakers, subject matter experts, civil society, and the people who use its products. The updated policy takes effect November 12, 2026.