Anthropic has expanded Claude’s memory capabilities so that the AI assistant can now share and carry over remembered information between its standard chatbot interface and its Cowork computer-use agent. The update, announced on Tuesday, means that when a user runs a task through Cowork in the cloud, the context Claude has gathered from prior chat conversations is available there, and details learned during Cowork sessions will also apply in future chats. According to the company, this reduces the need for users to repeat themselves when switching between the two products. The change builds on a recent upgrade to Anthropic’s Claude for Chrome browser extension, which made Cowork available through that plugin, meaning side-panel conversations now also count toward the shared memory history.
In addition to the cross-platform memory sync, Anthropic is introducing a new privacy control for what the model stores. Users will now see an option labeled “include sensitive topics in memory” in Claude’s settings. By default, the assistant does not save personal or sensitive subject matter, including details about a user’s health, race, ethnicity, religious beliefs, politics, gender identity, and similar areas. If a user enables that toggle, Claude will begin adding such topics to its memory, and each time it does so, the user will receive a notification. Anthropic stated that the change reflects the idea that what one person considers sensitive, another may find useful for the assistant to remember, though certain categories remain off-limits regardless of the setting.
Even with the sensitive topics feature turned on, Claude will not store specific high-risk data such as social security numbers, criminal history, or immigration status. Users can also edit or delete any individual memory entry at any time. The company clarified that enabling the new option will not cause Claude to retroactively pull in old conversations; only new exchanges will be considered for memory storage. This adds to earlier controls introduced over recent weeks that let users view memories organized by topic and edit the specifics of each topic from the settings menu.
The memory functionality is enabled by default for all account tiers, including Free, Pro, and Max, and is accessible on desktop and mobile. For users on Android or iOS, the latest version of the Claude app is required to see the new options. This update is part of Anthropic’s ongoing effort to make its AI tools more persistent and less fragmented in how they recall user interactions, particularly as Cowork becomes a more integrated part of the company’s product lineup for automating routine tasks. The move also gives US-based users more direct control over privacy boundaries in a consumer AI market where memory features have drawn scrutiny.
More AI news from TechManNews.
