Anthropic Usage Policy

Anthropic's Usage Policy (effective September 15, 2025) governs all users who submit inputs to Anthropic's products and services, including via authorized resellers and passthrough access, establishing binding behavioral standards as a condition of service access. The policy creates significant obligations through three tiered categories: Universal Usage Standards (applicable to all users), High-Risk Use Case Requirements (for elevated-risk consumer-facing deployments), and Additional Use Case Guidelines covering chatbots, minors, agentic use, and Model Context Protocol servers. Notably, the policy explicitly reserves Anthropic's right to modify government customer contracts to tailor use restrictions, and mandates reporting of CSAM and minor coercion to relevant authorities — an affirmative disclosure obligation that goes beyond standard platform AUPs. The policy engages the FTC Act Section 5 (unfair/deceptive practices), COPPA (children's online safety), EU AI Act (prohibited AI practices including manipulation and social scoring), GDPR Art. 6 (lawful basis for processing), and CFAA (computer fraud and abuse prohibitions). Compliance teams deploying Claude via API must assess whether their specific use case triggers High-Risk Requirements, and operators building agentic or MCP-based products face heightened obligations including human oversight and minimal footprint principles.

(1) REGULATORY EXPOSURE: This policy engages multiple regulatory frameworks simultaneously. The children's safety provisions implicate COPPA (15 U.S.C. §6501 et seq.), enforced by the FTC, and state-level equivalents. The prohibition on impersonation and psychologically manipulative content engages…