To the maximum extent permitted by law, Anthropic's total liability for any claims arising out of or relating to these Terms or our Services is limited to the greater of $100 or the amount you paid us in the last 12 months for the Services that gave rise to the claim.
This cap severely limits your financial recovery from Anthropic even if you suffer significant harm, such as from a data breach, incorrect AI output that causes professional or financial damage, or an unauthorized account action.
Anthropic's Consumer Terms permit the company to use your conversations, inputs, and outputs to train AI models by default, with a limited opt-out that does not apply when you provide feedback or when your content is flagged for safety review. Subscriptions auto-renew and fees are non-refundable upon cancellation, meaning you could be charged for a full billing period even if you cancel. You can opt out of conversation data being used for model training by navigating to your Claude.ai account settings and disabling the training opt-out toggle.