Anthropic has updated Claude’s memory system so information learned in chat can also carry into Claude Cowork, and vice versa. The change is designed to reduce the need for users to repeatedly explain the same project details when moving between conversation and task-based work.
Anthropic announced the update on Tuesday in an official Claude product post. The company also now lets users view everything Claude has stored in memory, organized by topic, and edit or delete individual entries.
Claude Now Shares Memory Across Chat and Cowork
Previously, information developed over multiple Claude chats did not automatically carry into Cowork. Users could therefore brainstorm a project in chat and then have to provide the same context again when asking Cowork to act on it.
With the new system, Cowork can use details Claude already remembers from chats, while information learned during Cowork sessions can also carry back into future conversations. Anthropic gives examples such as Cowork already knowing a manager’s preferences when drafting an update, or remembering a conference’s headcount, city, and speakers after those details were discussed earlier in chat.
Claude will also add topics to memory while a conversation is still happening instead of waiting until the end to create a summary. Anthropic says this allows updated details, such as a changed project deadline, to become available across Claude more quickly.
Users Can View and Edit Stored Memories
Claude now shows its retained information as a collection of files under Topics in Memory settings. Users can read, edit, or delete individual topics, as well as pause or reset memory entirely.
Anthropic says changing one stored detail also affects future chats and Cowork tasks that use that memory. The company described the update as a way to give users more direct control over what Claude retains.
Sensitive Topics Remain Off by Default
By default, Claude does not save sensitive topics including health information, race, ethnicity, religious beliefs, politics, or gender identity. Users can choose to enable an “include sensitive topics in memory” setting, and Claude will display a notice whenever it saves information from one of those categories.
Some information will not be stored even when that option is enabled. Anthropic says Claude will not retain government-issued identification numbers, Social Security numbers, criminal history, immigration status, or information that violates its Acceptable Use Policy.
The memory feature is enabled by default for Free, Pro, and Max users across web, desktop, and mobile. iOS and Android users need the latest version of the Claude app to access the newest changes, while Team and Enterprise administrators control whether memory is available within their organizations.
Featured image credits: SlideTeam
For more stories like it, click the +Follow button at the top of this page to follow us.