Claude Gets Memory — But Only for Select Paid Users
Anthropic is slowly rolling out one of the most requested AI features: memory. This week, the company announced that its Claude AI model can now “remember you and your team’s projects and preferences.” But there’s a catch — the feature is only available to users on the Team and Enterprise plans, leaving out free users and even those paying for individual Pro or Max subscriptions.
The update comes just days after Anthropic revealed new productivity tools that allow Claude to generate spreadsheets and presentation decks. Now, with memory, Claude can retain context across sessions, remembering client details, project histories, and even workflow patterns. If you use Claude’s Projects feature — essentially self-contained workspaces for different tasks — the AI will build a distinct memory for each, making it easier to pick up exactly where you left off.
Anthropic also highlighted new safety considerations around memory. The company says the feature is designed for work contexts and avoids storing sensitive or personal conversations, though it hasn’t gone into detail about the safeguards in place. For now, Anthropic says it will continue to monitor and test the feature before expanding it more widely.
For those with access, Claude memory can be enabled directly in app settings, where users can review what the AI remembers, edit or refine its “memory summaries,” and even delete certain entries. Team admins have the ability to switch off memory across their entire organization. To add flexibility, Anthropic also allows users to import memory from other AI systems or export Claude’s memory to different platforms — a nod to interoperability in the competitive AI space.
Notably, Anthropic rolled out another new feature for all users: incognito chats. Much like Chrome’s private browsing mode, these chats don’t appear in your conversation history and aren’t saved to memory — giving users more control over what persists and what doesn’t.
The timing of the rollout is strategic. Rivals like OpenAI’s ChatGPT and Google’s Gemini have had memory for months, with steadily expanding capabilities. Claude’s new functionality signals Anthropic’s determination to catch up while catering to business users first, where context retention can have the most impact.
With the feature still limited to enterprise customers, the big question is when — or if — Anthropic will bring memory to its broader user base. Until then, only teams with paid plans will get to experience a Claude that can finally remember.