One of the most annoying parts of working with an AI assistant is having to explain yourself all over again. Claude tries to fix it. Starting today, the memory used by Claude in chat is also available in Claude Cowork, meaning that the context built up over months of conversation can follow you when you task Claude.
This means that Claude can learn about your projects, priorities and preferences without needing another lengthy briefing. Explain in the chat how your team defines its metrics, for example, and Cowork can use that information to create the next quarterly business review. Discuss a conference in chat, and Cowork can provide details like city, speakers, and headcount as it works on logistics.
Claude now remembers the way you speak
There’s another change here that’s arguably just as important: Claude’s memory now refreshes while he’s chatting instead of waiting until the conversation is over to summarize it. Mention that the project deadline has moved to September, and Claude can carry that information into his next conversation without having to specifically tell him to remember.
And crucially, users can see what Claude actually remembers. Saved memories appear as individual files organized by topic that users can read, edit, or delete. Fix something once and the change will apply to future conversations and Cowork tasks.
You can have a say in what Claude remembers
By default, Claude does not save sensitive topics such as health, race, ethnicity, religious beliefs, politics, or gender identity. But users who want Claude to remember these details can now enable “memory sensitive topics”. When Claude posts something that affects one of these topics, users will see a notification and the setting can be turned off at any time.
What’s more, there are still categories that Claude doesn’t store even when sensitive memory is enabled, including sensitive ID numbers, criminal history, and immigration status. Also, the memory can be paused or resumed whenever users want.
The larger idea is pretty simple: Claude tries to make memory feel less like a hidden AI trick and more like a personal workspace that you actually control. And by having that memory follow you from chat to Cowork and back, Anthropic eliminates one of the biggest annoyances of working with AI: having to say the same thing 47 times.
\
