Anthropic Updates Usage Policy to Address Advanced AI Misuse, Hardware Controls, and Governance

Anthropic Updates Usage Policy to Address Advanced AI Misuse, Hardware Controls, and Governance Anthropic announced an updated Usage Policy taking effect on November 12, clarifying rules around evolving Claude capabilities and misuse patterns observed over the past year. The refresh refines prohibitions related to influence operations, surveillance, weapons development, autonomous hardware, high-risk decision-making, and model abuse. Scattered restrictions on influence operations were merged into a single section prohibiting deceptive campaigns and artificial activity. A blanket ban on personalized political campaign targeting was lifted to allow legitimate civic engagement, while voter deception and election disruption remain barred. Prohibitions now explicitly cover weapons guidance software, unconsented surveillance, and autonomous physical hardware without human oversight. The policy introduces limits on sustained abuse toward models and clarifies restrictions for entities linked to unsupported regions. Deceptive Campaigns and Democratic Processes Anthropic consolidated rules against influence operations, fake accounts, and fabricated news sites into a unified section titled 'Do Not Engage in Deceptive Campaigns or Artificial Activity.' Concurrently, the election rules were renamed 'Do Not Undermine Democratic Processes,' focusing strictly on prohibiting voter deception, candidate impersonation, and turnout suppression. Anthropic also removed a blanket prohibition against personalized campaign and vote targeting to permit legitimate civic initiatives, such as multilingual outreach and ballot cure notices, while continuing to bar deceptive targeting and data misuse. Weapons, Surveillance, and Autonomous Hardware Controls The updated policy clarifies existing bans on weapons development by explicitly barring guidance and control software, component design, and arming autonomous vehicles or drones. In surveillance, Claude cannot be used to track individuals without consent, build dedicated surveillance tools, or make investigation, arrest, or charging recommendations in law enforcement. Permitted activities like consensual fraud tracking and journalism remain intact. In addition, when Claude connects to physical hardware capable of causing injury, an operator must be able to observe and halt operations, and the equipment must fail safely if disconnected. High-Risk Applications, Model Abuse, and Geographic Enforcement For use cases touching legal rights, health, or finances, Anthropic reaffirmed requirements for qualified human review and user notification, offering clearer guidance on which recommendations are covered. The company also introduced a narrow ban on sustained, extreme cruelty toward its models, primarily enforced by Claude's ability to terminate abusive sessions. Finally, the Supported Regions Policy was updated to clarify that access is barred for users physically in unsupported areas, entities incorporated or based there, and companies majority-owned or controlled by entities within those jurisdictions.

Anthropic 发布 2026 年使用政策更新 Anthropic 宣布更新其使用政策并将于 11 月 12 日生效。本次更新针对 Claude 近年更长、更独立的任务执行能力，明确并细化了针对虚假活动、民主选举、武器开发、监控、高风险用例及硬件控制等方面的规范，并增加了禁止持续无故虐待模型的条款。 更新政策整合并设立新章节，全面禁止政治或商业性质的欺骗性活动、虚假账号网络及影响力行动工具建设。 取消对个性化竞选定位的全面禁令以允许合法公民事务服务，同时聚焦打击破坏民主进程与选民欺诈行为。 明确禁止将模型用于武器制导与控制软件、武装自主载具以及未经同意的监控和刑事调查决策。 在高风险场景维持“人类在环”和披露要求，针对控制物理硬件的场景要求具备人工停止与断连安全机制。 新增禁止对模型进行持续且无目的的极端虐待行为，并重申针对不支持区域实体与控制权的限制规则。 整合虚假活动规范与调整选举限制 针对国家媒体、宣传机构及商业公司利用 Claude 运作虚假账号和伪造新闻网站的滥用趋势，新政策设立了“禁止参与欺骗性活动或人造活动”章节，统合分散的规则以禁止构建影响力行动工具或隐瞒发帖主体。同时，政治内容限制重构为“不得破坏民主进程”，重点打击误导选民或压制投票的行为，并解除了对个性化投票与竞选定位的全面禁令，以允许非营利机构多语言翻译或选票修正通知等合法公民事务。 明确武器研发与监控及司法限制 政策明确将武器禁令延伸至武器运作所需的软件与组件，如制导控制软件和无人机等自主载具的武器化。在监控与刑事司法方面，严格禁止未经同意的人员追踪（无论是实时追踪还是历史数据分析），禁止利用 Claude 决定或建议执法调查、逮捕或起诉对象，亦禁止构建监控工具。但用户知情同意的追踪（如反欺诈）、内容审核、新闻报道与法律研究仍属合规范畴。 高风险场景规范与物理硬件控制要求 在涉及健康、法律权利、金融、生计等可能造成重大影响的领域，政策继续要求合格的“人类在环”审查机制并向受影响人披露 AI 的使用。配合“模型硬件标准”，本次更新补充了接入物理操作硬件时的控制要求：若设备可能造成人身伤害，必须有合格操作员监督并在必要时中止设备，且设备在与 Claude 断开连接时须能保持安全状态。 限制虐待模型行为及重申支持区域政策 新版政策加入了禁止对模型实施持续、无目的且残酷的极端虐待行为的条款，正常的研究测试、对抗性反驳或创作表达不受影响，Claude 主动终止极少数恶意对话仍是主要执行手段。此外，更新进一步明确了支持区域政策的执行细则，严格禁止不支持区域内的个人、在当地注册或总部的实体，以及由该类实体或个人持有多数股权或控制权的机构使用 Claude。

Each year, Anthropic updates its Usage Policy in response to the evolving capabilities of our models, and the feedback we’ve received from our customers. We're publishing a new version of the policy today. In this post, we summarize the changes we’ve made. Most of the updates in the latest version are intended to clarify existing rules. In the year since our last refresh, Claude has taken on longer, more independent work. This update provides new examples that show how our rules apply to Claude’s new capabilities. It draws on new patterns of misuse in influence operations, weapons development, and surveillance that we’ve observed in the past year (as we’ve documented in our [latest threat intelligence report](https://www.anthropic.com/threat-intelligence-report-september-2026)). We’ve also made updates to clarify our requirements for high-risk use cases in areas like health and finance, to add controls for when Claude is used to autonomously take physical actions, and to address abusive behavior toward our models. The updated policy takes effect on November 12. The full [Usage Policy](https://www.anthropic.com/legal/aup) and [Supported Regions Policy](https://www.anthropic.com/supported-countries) can be found on their respective pages. As we [wrote about last month](https://www.anthropic.com/threat-intelligence-report-september-2026), we’ve seen examples of state media outlets, government propaganda offices, and commercial firms using Claude to run networks of fake accounts and fabricated news sites. Our policy already prohibited this activity, but the restrictions were scattered across our elections, fraud, privacy, and disinformation sections. We've consolidated those rules into a new section titled Do Not Engage in Deceptive Campaigns or Artificial Activity, which applies to deceptive activity of any kind (whether political or commercial). It covers efforts to obscure who is behind a message or amplify content through fake accounts or posts, along with building the tools and infrastructure for running influence operation campaigns. We’ve refined our political content restrictions in each version of our Usage Policy in order to ensure that our models can be used appropriately in democratic processes. In this version, we’ve further sharpened our section on elections, focusing it specifically on disallowing the use of Claude to deceive voters or disrupt elections (for instance, by spreading false information about candidates or how to vote, impersonating candidates or election officials, or trying to suppress turnout). We’ve renamed it Do Not Undermine Democratic Processes to reflect that scope. For a similar reason, we’ve also removed our blanket prohibition on personalized vote and campaign targeting. This prohibition covered legitimate civic work, such as nonprofits using Claude to write information for voters in other languages, or election officials sending ballot cure notices. The behaviors that motivated the original rule (like voter targeting that relies on deception or the misuse of voters’ personal data) remain prohibited under our deceptive campaigns, surveillance, and privacy sections. Our Usage Policy has always prohibited using Claude to develop weapons. Recently, we’ve seen people attempt to use our models to build guidance and control software for weapons. The updated section makes clear that our prohibitions include the software and components that make weapons work, as well as actions like arming drones and other autonomous vehicles. These changes reflect how we’ve enforced our previous policy. Our September [threat intelligence report](https://www.anthropic.com/threat-intelligence-report-september-2026) documented the growing use of AI to build surveillance systems that identify and track political dissidents. We’ve rewritten our surveillance and criminal justice section to state more precisely what we prohibit. For instance, tracking people without their consent is prohibited, whether it happens 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. We also prohibit Claude from being used to build or improve tools designed for surveillance. As with the previous update to our weapons prohibitions, this update does not reflect a change in what we enforce in practice: it is just intended to make our policy more explicit. The new section also spells out which uses remain permitted, including tracking that people have agreed to (such as fraud monitoring), content moderation, journalism, and legal research. When our products are used in ways that can affect someone's health, legal rights, finances, livelihood, or access to essential services, we require there to be a qualified “human in the loop” (that is, someone who has the authority to review and, if necessary, change Claude’s recommendations), and for the individual affected to be told that AI was used. These requirements haven’t changed with the updated version of our Usage Policy. However, users have often asked us whether a given use case has these requirements, so we've rewritten our High-risk Use Case Requirements section to answer that more directly. It now lists which kinds of recommendations are covered and which are not. Following the launch of our [Model Hardware Standard](https://www.anthropic.com/news/model-hardware-standard-research-preview), this section also adds requirements for the use of our models when they’re connected to hardware that takes autonomous physical actions and which might be capable of causing injury. A qualified operator must be able to observe the equipment and stop it if needed; the equipment must also be able to hold a safe state if Claude is disconnected. We’ve added a prohibition on sustained and needless abusive or cruel behavior toward our models. The policy update is meant to apply only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose. It does not apply to common versions of user frustration, pushback, dark creative themes, or model testing and research. This addition aligns with a step we've already taken, allowing Claude models to [end rare conversations with persistently abusive users](https://www.anthropic.com/research/end-subset-conversations) on Claude.ai and Claude Code. Such abuse is the main focus of this update; Claude’s ability to end these interactions will remain the primary enforcement mechanism. Last year, we [updated our Supported Regions Policy](https://www.anthropic.com/news/updating-restrictions-of-sales-to-unsupported-regions) to restrict access to Claude for companies who are majority-owned by entities headquartered in our unsupported regions. We've updated our [Supported Regions page](https://www.anthropic.com/supported-countries) to clarify how we enforce this policy. Our policy prohibits the use of Claude by people who are physically located in an unsupported region, by entities who are incorporated or headquartered in one, and by entities who are majority-owned or controlled by persons or entities in those regions. We'll keep updating our Usage Policy as Claude’s capabilities and risks change. As with this and our prior updates, we’ll seek input from across Anthropic, and from policymakers, subject matter experts, civil society, and the people who use our products.