Claude Now Remembers Context Across Chats and Cowork

The update

Claude’s memory feature now spans both individual chats and Anthropic’s new Cowork workspace. Previously, context was often fragmented between sessions. With this change, conversations in one area become available in the other, so users do not need to re-add context when switching tools. The AI can now also update its memory in real time during a chat, recalling specific details without explicit prompting.

Why it matters

This unification creates a more persistent and personalized experience. By maintaining context across tools, Claude can better assist with ongoing projects and recall preferences on writing style or other details. The update also introduces new controls, allowing users to view, edit, or delete saved topics and manage what the AI stores.

What to watch

Users will see notifications when the AI adds sensitive topics to memory, which can be manually enabled. Anthropic states that certain sensitive information, such as government IDs or financial account numbers, will not be stored regardless of settings. The company is also rolling out the feature across web, desktop, and mobile platforms.

Sources

• cnet.com — Context about unifying memory across Claude and Cowork, real-time updates, and sensitive topic controls.
• engadget.com — Details on the fragmentation of previous memory and the new option to include sensitive topics.