Anthropic has released its annual update to its Usage Policy, a revision intended largely to clarify existing rules in light of the evolving capabilities of its Claude models. The company said the new version responds to feedback from customers and to new patterns of misuse observed over the past year, including in influence operations, weapons development, and surveillance. The updated policy takes effect on November 12.
A central change consolidates scattered restrictions on deceptive activity into a new section titled Do Not Engage in Deceptive Campaigns or Artificial Activity. This section applies to deceptive activity of any kind, whether political or commercial, and covers efforts to obscure who is behind a message, amplify content through fake accounts or posts, and build tools or infrastructure for influence operation campaigns.
Anthropic also refocused its elections section, renaming it Do Not Undermine Democratic Processes. The revised section specifically prohibits using Claude to deceive voters or disrupt elections, such as by spreading false information about candidates or voting procedures, impersonating candidates or election officials, or suppressing turnout. At the same time, the company removed a blanket prohibition on personalized vote and campaign targeting, which it said had covered legitimate civic work like nonprofits writing voter information in other languages or election officials sending ballot cure notices. Behaviors that motivated the original rule remain prohibited under other sections.
The update also clarifies prohibitions on weapons, making explicit that they include the software and components that make weapons work, as well as actions like arming drones and other autonomous vehicles. Similarly, the surveillance and criminal justice section was rewritten to state more precisely what is prohibited, including tracking people without consent and using Claude to decide or recommend who to investigate, arrest, or charge. Permitted uses such as consensual tracking, content moderation, journalism, and legal research are spelled out.
For high-risk use cases in areas like health and finance, the policy continues to require a qualified human in the loop and disclosure that AI was used, but the section has been rewritten to answer user questions more directly. Following the launch of Anthropic's Model Hardware Standard, new requirements apply when models are connected to hardware that takes autonomous physical actions: a qualified operator must be able to observe and stop the equipment, and the equipment must be able to hold a safe state if Claude is disconnected.
Anthropic added a prohibition on sustained and needless abusive or cruel behavior toward its models, meant to apply only in extreme cases and not to common frustration, pushback, dark creative themes, or model testing. This aligns with an existing step allowing Claude models to end rare conversations with persistently abusive users on Claude.ai and Claude Code. The company also clarified its Supported Regions Policy, reiterating restrictions on access from unsupported regions and for entities incorporated, headquartered, or majority-owned there.