Anthropic's official account @claudeai posted on X on August 25, 2026, announcing: "Claude now has one memory across chat and Claude Cowork, and you decide what's in it. Hand Cowork a task and it starts from what Claude already knows from your chats: the project you talked through, your manager's preferences, or the client from last quarter."
In short, Claude now shares a single memory (memory — a feature that recalls prior conversations to inform later interactions) across chat and Claude Cowork, and users decide what goes into it.
What the Announcement Says
According to the post, when a user hands a task to Claude Cowork, Claude starts from what it already knows from earlier chats, without needing the context repeated. The post cites examples such as a project previously discussed, a manager's preferences, or details about a client from last quarter.
What It Means for Users
This could reduce the need to re-explain the same background every time users switch between chat and Cowork. The post also emphasizes that users decide what's stored in memory, underscoring user control over what the memory contains.
Not Covered in the Official Announcement
The announcement does not specify the rollout scope (all users or specific plans), how long memory is retained, or detailed management and deletion controls. These details were not included in the official announcement.
For the latest details, please check Anthropic's official channels.
Related articles
- Claude Chat Projects vs. Cowork Projects — What Carries Over (Tested)
- Claude Chat vs. Cowork — Which One for This Task? A Simple Decision Guide
- Claude Projects, Artifacts & Memory: What They Are and How to Use Them
- What Is Claude Cowork? Claude's File-Working Desktop Mode
- Claude Memory — How to Make Claude Remember and Manage It