Anthropic Unifies Claude Chat and Cowork Memory for Persistent Context

Anthropic Unifies Claude Chat and Cowork Memory for Persistent Context

Anthropic unified the memory systems for Claude chat and Claude Cowork on Tuesday, August 25, enabling Claude to carry conversational context between both experiences by default. This update eliminates the need to re-brief Cowork, enhancing continuity for users across Free, Pro, and Max plans on web, desktop, and mobile. For broader context, explore our AI News.

Enhanced Continuity Across Claude Platforms

The core of this update is the merging of memory systems that previously operated independently for Claude chat and Claude Cowork. With this change, Claude can now carry context from one experience to the other. For instance, a conversation initiated in Claude chat can be seamlessly continued in Cowork, with the AI retaining knowledge of previous interactions and shared information. This feature is designed to streamline workflows and improve the overall user experience by making Claude a more consistent and knowledgeable collaborator.

Users now have direct control over their stored memories. The system allows for incremental memory addition during a conversation, meaning Claude learns and stores information as the interaction progresses. Furthermore, users can read, edit, or delete any stored memory, providing flexibility and control over the data Claude retains. This level of user management is a key aspect of the unified memory system.

Availability and Data Privacy Considerations

The unified memory feature is now enabled by default across all Claude plans: Free, Pro, and Max. It is accessible on the web, desktop applications, and mobile devices. Users on iOS and Android platforms will need to update their respective applications to access this new functionality.

Anthropic has also addressed data privacy concerns related to this persistent memory. The company states that sensitive data, such as health information, ethnicity, or political views, will be excluded from storage unless users explicitly opt in. Additionally, certain highly sensitive information, including government IDs, Social Security numbers, criminal history, and immigration status, will not be stored by Claude under any circumstances. This approach aims to balance enhanced functionality with user privacy and data security.

Implications for AI Collaboration

This unification of memory systems positions Claude as a more integrated conversational AI, potentially simplifying complex tasks that span multiple interactions or require long-term context. Unlike earlier models that might require repeated introductions to a topic, Claude's enhanced memory allows for a more natural and efficient collaborative experience. This development aligns with broader industry trends towards creating more persistent and context-aware AI assistants, similar to how other platforms are exploring agent-like features to maintain user state and project continuity.

Conclusion

Anthropic's integration of Claude chat and Cowork memory marks a significant step towards creating a more cohesive and intelligent AI assistant. By enabling persistent context across its platforms, Claude aims to offer a more efficient and user-friendly experience for a wide range of tasks. Users are encouraged to update their mobile applications to access this feature and explore its capabilities for enhanced collaboration.

Sources

• [Feature Request] Unified user state across Cowork and Claude chat — shared memory, files, skills, connectors · Issue #55842 · anthropics/claude-code · GitHub
• When AI builds itself
• Introducing Claude
• Use Claude’s chat search and memory to build on previous context | Anthropic Help Center
• Release notes | Anthropic Help Center - Claude Support