Anthropic Privacy Policy

This document explains how Anthropic collects, uses, and shares your personal information when you use its Services. By default, Anthropic uses your conversations with Claude to train its AI—you can turn this off in your account settings, though safety-flagged or self-reported content can still be used. If you give Claude permission to act on your behalf (such as sending messages or accessing other services), your data may be sent to those outside services, and Anthropic is not responsible for how those services handle your data.

This privacy policy establishes Anthropic's data collection practices, usage rights, sharing permissions, and user controls with respect to personal data processed through its Services. Anthropic collects biometric data—including government-issued identity document images, photos or videos, and facial geometry templates—for identity verification. By default, Anthropic may use user Inputs and Outputs to train and improve its AI models, subject to a user-exercisable opt-out through account settings; that opt-out is overridden when content is flagged for safety review or explicitly reported by the user. Anthropic may share personal data with government authorities and law enforcement on a good-faith belief of legal necessity, and may transmit user Inputs, Outputs, and instructions to Third-Party Services to perform agentic actions on the user's behalf, while explicitly disclaiming responsibility for those services' data practices. Anthropic states it does not engage in consequential automated decision-making based solely on automated processing or profiling.

As a user, your conversation data—including your inputs and Claude's responses—is used by default to train Anthropic's AI models. You can stop this by opting out through your account settings, though content flagged for safety review or content you yourself report to Anthropic remains subject to use for model improvement regardless of your opt-out. Anthropic also collects biometric data including facial geometry templates when you verify your identity. If you grant Claude permissions to act outside the platform—such as sending communications or interacting with third-party services—your inputs, outputs, and instructions may be transmitted to those third parties, and Anthropic does not control or take responsibility for how those third parties handle your data. The concrete action available to you is: opt out of model training through your account settings.