Anthropic Introduces Persistent Memory and Incognito Chat for Claude

AI safety firm Anthropic has launched a persistent memory feature for its Claude chatbot, enhancing productivity tools and privacy controls for Team and Enterprise users. Starting September 12, Claude AI will automatically remember project details and user preferences across conversations, removing the need to repeatedly explain context.

This significant upgrade is available exclusively to Claude’s Team and Enterprise subscribers. Anthropic has also introduced an 'Incognito chat' mode for all users, ensuring conversations remain private and outside the AI’s memory. This move positions Claude to better compete with rivals by focusing on enterprise-grade productivity.

The new memory system is a true persistent memory that actively learns from interactions, marking a shift from a manual search tool to an automatic recall engine. This feature addresses a key concern for businesses handling sensitive information.

Anthropic is directly targeting this productivity pain point. 'Claude now remembers your past conversations, so you can seamlessly continue projects, reference previous discussions, and build on your ideas without starting from scratch every time,' the company states.

The new Incognito mode provides a 'clean slate' for sensitive brainstorming or confidential discussions. This update is a direct response to the escalating AI memory arms race.

This feature is part of a broader push by Anthropic to build a suite of tools for the professional market. While persistent memory introduces significant security challenges, Anthropic aims to build trust with a professional user base through advanced capabilities and transparent controls.