Claude.ai Terms of Service

This document governs individual consumer use of Claude.ai, Claude Pro, and associated Anthropic products and services, establishing a binding contract between users and Anthropic, PBC that incorporates the Acceptable Use Policy and Privacy Policy by reference. The terms authorize Anthropic to use user-submitted inputs and outputs (collectively 'Materials') to train AI models unless users opt out through account settings, though the agreement states that opt-out does not apply when users provide Feedback or when Materials are flagged for safety review; the terms also assign Anthropic-held output rights to users subject to compliance, and impose mandatory binding arbitration with a class action waiver for US users. Notably, the training data opt-out contains carve-outs that may render it narrower in practice than consumers expect, and the arbitration clause includes a 30-day opt-out window from the date of first account creation or the effective date of these terms, which is a common but consequential provision; the agreement also disclaims all warranties and caps liability at the greater of fees paid in the prior 12 months or $100, which is standard but significant given the AI context. The terms engage GDPR and UK GDPR for European and UK users respectively, CCPA for California residents, COPPA implicitly through the age-18 minimum, and the EU AI Act may require evaluation given the nature of the services; the mandatory arbitration and class action waiver provisions engage FTC consumer protection authority and may face enforceability limitations in certain jurisdictions including EU member states.