Anthropic has introduced a new Memory feature for its Claude AI, designed to make teamwork more seamless and efficient. Here are the key takeaways:
-
Context that Stays with You
Claude can now remember project details, client information, workflows, and team preferences across conversations. This means less time re-explaining and more time moving work forward. -
Project-Based Boundaries
Memory isn’t a one-size-fits-all log. Details are separated by projects, so information from a product launch doesn’t accidentally show up in a client conversation. -
Full User Control
Teams can see and edit what Claude remembers. You can update or delete parts of the memory and even tell Claude to ignore certain details altogether. -
Incognito Chat for Sensitive Work
For one-off or confidential discussions, users can switch to Incognito chat. Nothing from these sessions is stored in memory or conversation history. -
Enterprise Admin Options
The feature is rolling out to Team and Enterprise plans first. Admins have the ability to disable memory entirely at an organizational level if they choose.
Summary
Anthropic’s new Memory feature gives Claude the ability to retain project-specific context, helping teams save time and reduce repetition. With incognito options, admin controls, and privacy safeguards, the update aims to balance smarter AI assistance with user trust. It positions Claude as a serious contender in the growing competition over AI memory systems.
