Anthropic's 2026 Usage Policy: abuse of Claude, election rules, and physical actions

Anthropic has published a new version of its Usage Policy, the rules for how Claude may be used across its apps and API. The 8 October 2026 announcement says the updated policy “takes effect on November 12.” Anthropic describes most of the changes as clarifications prompted by Claude doing “longer, more independent work,” and by misuse it has documented in the past year.

A new section on deceptive campaigns

Rules against fake accounts and fabricated news sites were spread across the election, fraud, privacy and disinformation sections. Anthropic has consolidated them into one section, “Do Not Engage in Deceptive Campaigns or Artificial Activity.” It covers hiding who is behind a message, amplifying content through fake accounts or posts, and building infrastructure for influence operations, whether the aim is political or commercial.

A narrower elections section

The elections section is renamed “Do Not Undermine Democratic Processes.” It now focuses on deceiving voters or disrupting elections: false information about candidates or voting procedures, impersonating candidates or officials, and suppressing turnout.

Anthropic also removed its blanket prohibition on personalized vote and campaign targeting. It says that rule blocked legitimate civic work, such as nonprofits writing voter information in other languages or election officials sending ballot cure notices. Targeting that relies on deception or misuse of voters’ personal data, it says, remains prohibited under the deceptive-campaigns, surveillance and privacy sections.

Weapons and surveillance, spelled out

The weapons section now states that the ban includes “the software and components that make weapons work,” including guidance and control software and arming drones or other autonomous vehicles. The surveillance and criminal justice section prohibits tracking people without consent, in real time or from stored data, and bars Claude from deciding or recommending “who to investigate, arrest, or charge.” Anthropic says both rewrites reflect how it already enforces the policy, and lists permitted uses such as consented fraud monitoring, content moderation, journalism and legal research.

High-risk uses and physical actions

Uses that affect health, legal rights, finances, livelihood or access to essential services still require a qualified human in the loop and disclosure to the affected person. The section has been rewritten to list which recommendations are covered.

New this year: when Claude is 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 hold a safe state if Claude disconnects.

Abuse toward the models

The update adds a prohibition on “sustained and needless abusive or cruel behavior toward our models.” Anthropic says it applies “only in extreme cases” of repeated cruelty with no discernible purpose, not to frustration, pushback, dark creative themes, or testing and research. Claude’s existing ability to end persistently abusive conversations “will remain the primary enforcement mechanism.”

Supported regions

Anthropic also clarified enforcement of its Supported Regions Policy: use is prohibited by people physically in an unsupported region, by entities incorporated or headquartered there, and by entities majority-owned or controlled from those regions.

What to do

If your product touches any of these areas, compare your use case with the full Usage Policy text before 12 November. The announcement is a summary; the policy page is what applies.