Anthropic adds Memory to Claude for Team and Enterprise users

Anthropic has announced a new memory feature for the Claude app, directly targeting users on Team and Enterprise plans. This memory capability is rolling out now, with Enterprise admins empowered to disable it as needed, while Incognito chat is being made available to all users regardless of plan. The feature is currently limited to these groups, with a phased approach planned before any wider rollout.

The memory function is designed to store and recall details about ongoing projects, client needs, and team processes, reducing the need for users to repeat information. Each project receives a distinct memory compartment, keeping conversations and sensitive information separate. Users have granular control over what Claude remembers and can view or edit memory summaries via settings. Technical options include toggles for memory generation and chat search, and users can import or export memory data for migration or backup. Incognito mode provides a way to conduct conversations without saving them, supporting privacy for sensitive discussions.

Compared to previous Claude versions, this feature introduces persistent context across chats, which has been a frequent request from enterprise and power users. Competing platforms like OpenAI’s ChatGPT and Google Gemini have implemented memory features, but Claude’s project-based separation and user-controlled memory editing set it apart. Early user feedback from pilots indicates improved workflow continuity and reduced repetition, while enterprise customers have highlighted the value of adjustable privacy and memory controls.

Anthropic, the company behind Claude, is known for its focus on responsible AI development and safety. This launch reflects their intent to support professional teams handling complex, multi-project workflows, reinforcing Claude’s position in the competitive business AI assistant market.

Source