What changed: In its annual Usage Policy refresh, Anthropic added a new section on deceptive campaigns and artificial activity, renamed and refocused its elections section on protecting democratic processes, sharpened weapons and surveillance prohibitions, clarified high-risk use case requirements, added controls for autonomous physical actions via connected hardware, and prohibited sustained abusive behavior toward its models.
Who is affected
The source states this covers businesses using Anthropic's products via Claude.ai, Claude Code, and the API, with particular relevance for high-risk use cases (health, legal, finance) and users connecting Claude to autonomous physical hardware.
What it means for your business
For businesses using Claude or the Anthropic API in their products or operations, this means complying with the updated policy starting November 12, including new oversight and safe-stop requirements for uses connected to autonomous physical hardware.
What to do
Businesses using Anthropic products should review the updated Usage Policy and Supported Regions Policy before November 12 and check compliance with the new requirements for autonomous physical hardware use and high-risk use cases.
When it takes effect
The updated policy takes effect on November 12.
Source: https://www.anthropic.com/news/2026-usage-policy-update
