Each year, Anthropic revises its Usage Policy to reflect the changing capabilities of its models and the feedback from customers. A new version is being published today, and this post summarizes what has changed.
Most of the changes clarify rules that already exist. Over the past year, Claude has taken on longer and more independent tasks. The update adds new examples showing how the rules apply to these capabilities. It also draws on new patterns of misuse in influence operations, weapons development, and surveillance that Anthropic has observed over the past year, as described in its latest threat intelligence report. Anthropic has also clarified requirements for high-risk use cases such as health and finance, added controls for cases where Claude autonomously takes physical actions, and addressed abusive behavior toward its models.
The updated policy takes effect on November 12. The full Usage Policy and Supported Regions Policy are available on their own pages.
A new section on deceptive activity
As described last month, Anthropic has seen state media outlets, government propaganda offices, and commercial firms use Claude to operate networks of fake accounts and fabricated news sites. The policy already banned this, but the rules were spread across the elections, fraud, privacy, and disinformation sections.
Anthropic has combined these rules into a new section, Do Not Engage in Deceptive Campaigns or Artificial Activity. It covers any deceptive activity, political or commercial. It addresses hiding who is behind a message, amplifying content through fake accounts or posts, and building tools and infrastructure for influence operations.
Refocusing the elections section
Anthropic has refined its political content restrictions with each version of the policy so that its models can be used appropriately in democratic processes. The elections section is now narrower. It focuses on banning the use of Claude to deceive voters or disrupt elections, for example by spreading false information about candidates or voting, impersonating candidates or election officials, or suppressing turnout. It has been renamed Do Not Undermine Democratic Processes.
Anthropic has also dropped its blanket ban on personalized vote and campaign targeting. That ban caught legitimate civic work, such as nonprofits writing voter information in other languages or election officials sending ballot cure notices. The behaviors that prompted the original rule, such as deceptive voter targeting or misuse of voters' personal data, stay prohibited under Anthropic's deceptive campaigns, surveillance, and privacy sections.
Clarifying weapons prohibitions
The policy has always banned using Claude to develop weapons. Anthropic has recently seen people try to use its models to build guidance and control software for weapons. The updated section states that the ban covers the software and components that make weapons function, and actions such as arming drones and other autonomous vehicles. These changes match how Anthropic has enforced the earlier policy.
More precise language on surveillance and law enforcement
The September threat intelligence report documented growing use of AI to build surveillance systems that identify and track political dissidents. Anthropic has rewritten its surveillance and criminal justice section to say more exactly what is banned. Tracking people without their consent is prohibited, whether in real time or through analysis of previously collected data. Claude cannot be used to decide or recommend whom to investigate, arrest, or charge in a law enforcement or criminal justice process. Building or improving tools designed for surveillance is also prohibited.
As with the weapons update, this does not change what Anthropic enforces in practice. It only makes the policy more explicit. The section also lists the uses that remain allowed, including tracking people have agreed to (such as fraud monitoring), content moderation, journalism, and legal research.
Clearer requirements for high-risk use cases
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" is required. This is someone with the authority to review and, if needed, change Claude's recommendations. The affected person must also be told that AI was used.
These requirements are unchanged. But users often asked whether a particular use case is subject to them, so Anthropic rewrote the High-risk Use Case Requirements section to answer more directly. It now lists which kinds of recommendations are covered and which are not.
Following the launch of the Model Hardware Standard, the section also adds requirements for using Anthropic's models connected to hardware that takes autonomous physical actions and could cause 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.
Addressing abusive behavior toward models
Anthropic has added a prohibition on sustained, needless abusive or cruel behavior toward its models. It is meant to apply only in extreme cases, where users repeatedly act cruelly toward the models with no discernible purpose. It does not cover ordinary user frustration, pushback, dark creative themes, or model testing and research.
This matches a step Anthropic has already taken: letting Claude models end rare conversations with persistently abusive users on Claude.ai and Claude Code. That is the main target of this update, and Claude's ability to end such interactions will remain the primary means of enforcement.
Clarifying the Supported Regions Policy
Last year, Anthropic updated its Supported Regions Policy to restrict Claude access for companies majority-owned by entities headquartered in unsupported regions. Anthropic has now updated its Supported Regions page to explain how the policy is enforced. The policy prohibits use of Claude by people physically located in an unsupported region, by entities incorporated or headquartered there, and by entities majority-owned or controlled by persons or entities in those regions.
Looking ahead
Anthropic will keep updating the Usage Policy as Claude's capabilities and risks change. As with this and earlier updates, it will seek input from across Anthropic, as well as from policymakers, subject matter experts, civil society, and the people who use its products.