Anthropic unifies Claude memory across Chat and Cowork
Anthropic has rolled out a single shared memory layer that links Claude Chat and Claude Cowork, letting users control what is stored rather than relying on siloed session context. Cowork tasks can now begin from information Claude already holds from prior conversations, such as active projects, manager preferences, or clients discussed in earlier quarters. The update also adds a dedicated Memory tab, splits remembered topics into separate categories, and introduces a Settings toggle for keeping memories about sensitive subjects. For operators and builders, the shift reduces repetitive briefing when moving from chat to agentic work, but Anthropic has not published detailed retention policies or migration timelines in the available announcement, so teams should treat stored context as user-governed until further documentation arrives.
Anthropic unifies Claude memory across Chat and Cowork
Anthropic rolled out one shared memory layer for Claude Chat and Claude Cowork, with users deciding what is stored. Cowork tasks can start from context Claude already knows from prior chats, including projects, preferences, and past clients.
Key takeaway
A unified Claude memory layer now lets Cowork reuse chat context users choose to store, bridging conversational and agentic workflows.
What happened
Anthropic rolled out one shared memory layer for Claude Chat and Claude Cowork, with users deciding what is stored rather than the product keeping separate memory silos for each surface.
Cowork tasks can start from context Claude already knows from prior chats, including projects, preferences, and past clients, and the rollout also adds a dedicated Memory tab, categorized topics, and a Settings toggle for sensitive-topic memories.
Evidence
Anthropic unified memory across Claude Chat and Claude Cowork with user control over stored content.
X · attributed
Anthropic rolled out one shared memory layer for Claude Chat and Claude Cowork, with users deciding what is stored.
Cowork can start tasks using prior chat context such as projects, preferences, and past clients.
X · attributed
Cowork tasks can start from context Claude already knows from prior chats, including projects, preferences, and past clients.
The memory update adds a dedicated Memory tab and splits remembered topics into categories.
X · attributed
Memory now has a dedicated tab in Claude
Users can toggle whether Claude keeps memories about sensitive topics in Settings.
X · attributed
a new toggle to keep memories about sensitive topics has been added to the Settings menu.
Why it matters
Teams running Claude in both chat and Cowork can reduce manual context re-entry, but compliance reviews must account for user-controlled retention of sensitive-topic memories.
Limits and uncertainties
The available packet is an Anthropic-owned X announcement and does not specify retention duration, enterprise controls, or rollout timing by plan or region.
Practical implications
Builders should design Cowork workflows assuming prior chat context may auto-populate tasks and verify users have reviewed what Claude is allowed to remember.
Operators should audit the new sensitive-topic memory toggle and category controls before relying on unified memory for regulated or client-confidential work.
What to watch
Whether Anthropic publishes detailed memory retention, deletion, and enterprise governance documentation beyond the initial announcement.
Original reporting: 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
