Anthropic has started implementing enhanced memory for Team and Enterprise users in the Claude chatbot. Now Claude automatically remembers details of previous conversations and considers them in new responses without additional prompts. This allows for taking into account user preferences, project context, and main priorities when working with the chatbot.
Claude’s memory extends to projects where users can create diagrams, website layouts, graphics, and other materials based on uploaded files. For each individual project, Claude maintains a separate memory. Users can view and edit saved data through the settings menu, as well as determine what Claude should pay attention to or ignore.
Anthropic emphasizes that the use of memory remains entirely voluntary. The feature is currently available only for Team and Enterprise, although previously paid users could manually ask Claude to remember past chats. Additionally, the company has launched “incognito chats” for all users, which Claude does not save or use in the future.
Similar context-saving features have already been implemented by OpenAI and Google in their chatbots. In Claude, they are primarily focused on work processes and team needs, such as “team processes” or “client requests.”
