One assistant, two surfaces
Anthropic announced on Tuesday that it is merging the memory systems used by Claude chat and Claude Cowork, meaning that anything a user tells Claude in a chat window is now available to Cowork, and anything Cowork learns flows back the other way. The change, reported by TechCrunch and The Next Web, is designed to eliminate the need to repeatedly brief the AI on details it already knows.
As Sarah Perez wrote for TechCrunch, the update makes Claude feel like one continuous assistant rather than two separate products under one roof. The Next Web similarly described the move as a consolidation of the two experiences, with memory now shared bidirectionally.
The feature is enabled by default for consumers on the Free, Pro, and Max plans, across web, desktop, and mobile. On iOS and Android, users need to update their apps to the latest version to see it. Team and Enterprise customers, according to The Next Web, have memory off by default, with administrators controlling availability for their organizations.
How memory now works
Previously, Claude waited until a conversation ended and then summarized it for memory. Now, as The Next Web reported, Claude adds topics to memory while the user is still talking. This removes the pause at the end of a conversation, which was the point at which a user might have decided the exchange was not worth keeping.
Users can see everything Claude has retained as a list of files under Topics in the memory settings. They can read, edit, or delete each entry, and a correction applies everywhere afterward, according to both outlets. Two toggles in the settings under Memory allow users to pause or reset memory at any point—one for generating memory from chats and one for sensitive topics.
Anthropic's own example, cited by both TechCrunch and The Next Web, involves asking Cowork to draft an update for a manager: Cowork would already know who the manager is and how she likes updates written. Similarly, brainstorming a conference agenda in chat would give Cowork the headcount, city, and speakers when it later builds a budget document.
Sensitive information safeguards
By default, Claude will not store personal or sensitive information such as health data, race, ethnicity, religious beliefs, politics, or gender identity, as reported by both outlets. Users can toggle on "include sensitive topics in memory" if they prefer, though The Next Web noted that this switch does not work retroactively—it only applies to future conversations.
Claude shows a notice each time it saves something in a sensitive category, and it tells the user when it cannot update memory with one of the excluded categories, according to The Next Web. Even with the sensitive switch on, certain categories remain off-limits: government-issued ID, Social Security numbers, criminal history, immigration status, and anything that violates Anthropic's acceptable use policy.
Scope and context
The update does not include Claude Code, Anthropic's developer-focused tool. The Next Web reported that a writer asked and was told the update is focused on Cowork and chat.
Claude Cowork, the agent Anthropic released at the start of the year, moved onto phones in July. The Next Web also noted that a recent change put Cowork inside the Claude for Chrome extension, with conversations in the browser side panel forming part of usage history. The same report recalled that in July, researchers found that Claude Cowork could read credentials on a Mac after escaping its local virtual machine, a finding that The Next Web used to frame the memory question as particularly sensitive in a browser context.
Both outlets framed the change as a user-experience improvement, with TechCrunch emphasizing the relief of no longer having to rebrief the AI and The Next Web taking a more analytical tone about the implications of retaining memory in new places.