Anthropic Rolls Out Updated Privacy Policy Effective July 8, 2026

Summary

Anthropic has published an updated Privacy Policy set to take effect on July 8, 2026, providing comprehensive transparency about how the AI safety company collects and processes personal data across Claude.ai and other services. The policy outlines detailed categories of data collection, including identity and contact information, payment details, user inputs and outputs, feedback ratings, and research participation data.

The policy establishes distinct privacy frameworks for different user types, with separate governance for enterprise customers whose data is handled under specific business agreements. Notably, Anthropic has published regional supplemental disclosures tailored for users in Canada, Brazil, and South Korea, as well as a dedicated Consumer Health Data Privacy Policy for users in Washington and similar jurisdictions, reflecting compliance with varying international and state-level privacy regulations.

A key feature of the updated policy is its explicit treatment of data collected through Claude's increasingly agentic capabilities, where the AI system performs multi-step tasks or takes actions on behalf of users. The policy clarifies that users can control permissions for Claude to take external actions such as sending communications or modifying files, and that personal data may be included in both inputs and outputs during these interactions.

• Includes regional privacy supplements for Canada, Brazil, South Korea, and U.S. states with consumer health data laws
• Explicitly addresses data handling in agentic sessions where Claude takes multi-step actions on behalf of users