Anthropic’s Claude can now automatically “remember” details from previous conversations without requiring users to prompt it each time. The enhanced memory feature, currently rolling out exclusively to Team and Enterprise subscribers, brings Claude to parity with OpenAI’s ChatGPT and Google’s Gemini, both of which already offer cross-chat memory features.
Claude will automatically incorporate user preferences, project context, and priorities into responses. The system focuses on work-related details like team processes and client needs. Unlike the manual memory prompting that launched last month for paid users, this upgrade operates automatically in the background.
The AI maintains separate memories for each project workspace, preventing context from overlapping between different initiatives. This integrates with Anthropic’s existing Projects tool, where users can generate diagrams, websites, and graphics from uploaded files.
Memory storage is completely optional. Users can view, edit, or delete stored memories through their settings menu and direct Claude to focus on or ignore specific details based on their needs.
Anthropic is also introducing incognito chats for all users, which won’t be saved to chat history or referenced in future interactions.
If you haven’t used the memory feature in ChatGPT or Gemini, you don’t know what you’re missing. (Feel free to ask your preferred provider about security and training, if that is a concern. Better to hear the facts directly from your AI provider than anyone else.) You should also check with your SecOps and Risk & Compliance peeps before entering any sensitive or corporate information in any context window – regardless of whether memory is on or off.
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it. This work was created with the assistance of various generative AI models.