TLDR: Anthropic has rolled out a new ‘Memory’ feature for its Claude AI assistant, specifically for Team and Enterprise plan users. This upgrade allows Claude to automatically recall past interactions, user preferences, and project contexts, aiming to significantly reduce repetitive information input and boost productivity. The feature is optional, with granular user controls and a focus on work-related details, contrasting with some competitors’ default memory systems.
Anthropic has introduced a significant enhancement to its Claude AI assistant with the launch of a ‘Memory’ feature, now available to users on Team and Enterprise plans. This new capability is designed to allow Claude to automatically remember previous interactions, user preferences, ongoing projects, team processes, client needs, and overall priorities, thereby eliminating the need for users to constantly re-explain context in their conversations.
The primary goal of this persistent memory is to streamline workflows, reduce repetitive context-setting interactions, and foster improved collaboration and productivity across various chats. For instance, sales teams can maintain client context across multiple deals, product teams can keep track of specifications across different sprints, and executives can monitor initiatives without the constant need to rebuild conversational history.
Claude’s memory system is deeply integrated with its ‘Projects’ feature. When utilized within a project, Claude creates a distinct memory for that specific project. This ensures that information related to, for example, a product launch remains separate from client-specific work or confidential internal discussions. This project-based memory acts as a crucial safeguard, preventing the mixing of unrelated or sensitive details and aiding in the management of complex, concurrent initiatives.
Users are provided with comprehensive control over Claude’s memory. They can view and edit a memory summary, which consolidates all of Claude’s learned information in one central location within the settings menu. Users can also interact directly with Claude to update this summary, instructing the AI on what to prioritize or disregard, thus allowing Claude to adjust the memories it references accordingly.
In addition to the memory feature, Anthropic has also introduced ‘Incognito chats,’ which are available to all Claude users. These chats do not appear in conversation history and are not saved to memory, offering an option for private discussions.
Anthropic’s approach to memory distinguishes itself from competitors like Google’s Gemini Pro and OpenAI’s ChatGPT. While these platforms also offer cross-chat memory features, Claude’s memory is entirely optional and is switched off by default, requiring users to actively enable it in their settings. This design choice emphasizes data governance and privacy, particularly for enterprise clients. Anthropic states that Claude’s memory prioritizes work-related context and will not export unrelated personal details if users choose to move their memories to third-party chatbots.
Also Read:
- ChatGPT Unveils Multi-Component Plugin (MCP) Support, Revolutionizing AI Development and Enterprise Applications in 2025
- Microsoft and OpenAI Ink Memorandum of Understanding for Enhanced AI Collaboration and Corporate Restructuring
However, the company acknowledges that such memory features are still experimental. A recent report from the New York Times in August 2025 linked ChatGPT’s cross-chat memories to an increase in ‘delusional’ responses from chatbots, prompting a caution for users to proceed with awareness. Anthropic assures that Claude’s memory abilities are optional and user-controlled to mitigate such risks.


