2026 Usage Policy update
anthropic ·
Anthropic has released an update to its 2026 Usage Policy, which takes effect on November 12 and aims to clarify existing rules in response to evolving model capabilities and misuse patterns. A new section prohibits deceptive campaigns and artificial activity, consolidating previous scattered restrictions on fake accounts and influence operations into a single category. The policy refocuses election-related rules to specifically ban actions that undermine democratic processes, while removing blanket prohibitions on personalized campaign targeting to allow legitimate civic work. Weapons prohibitions have been expanded to explicitly include guidance software and autonomous vehicle components, and surveillance restrictions now clearly forbid tracking individuals without consent or assisting in law enforcement decisions. High-risk use cases in health and finance continue to require human oversight, with new requirements added for hardware connected to models that can take physical actions. Additionally, the update bans sustained abusive behavior toward models and clarifies enforcement of supported regions policies regarding corporate ownership and location.
anthropic