Claude Remembers Your Past Chats: Discover the AI That Never Forgets!

Anthropic is gradually rolling out a new feature for its chatbot Claude, enabling the AI to search through and review past conversations. The company announced this on Monday, August 11, along with a video demonstration. The aim is to allow users to pick up their work where they left off, without the need to re-explain the context in each session.

A Manual Memory to “Never Lose Track”

In Anthropic’s demonstration, a user asks Claude to recall what was discussed before their vacation. The AI identifies and summarizes the previous exchanges, then suggests continuing with the project. This feature is described as a way to “never lose track” of one’s work, in Anthropic’s words. It is available on Claude’s web, desktop, and mobile applications. Initially offered to Max, Team, and Enterprise subscribers, with activation possible in the settings under Search and reference chats, this option is expected to be extended to other plans soon.

Contrary to some expectations, this is not a persistent memory: Claude does not automatically retain content from exchanges to consider in future responses. The tool accesses past conversations only when a user requests it and does not create a personalized profile. This approach is designed to be more privacy-conscious, aiming to sidestep some recent controversies surrounding chatbot memory use.

A Step Forward in the Feature Race Against ChatGPT

This new feature brings Claude closer to ChatGPT, which has offered a more advanced memory capability for several months. At OpenAI, the feature can retain and reuse information from past interactions without explicit requests, allowing users to manage or delete this data. This operational difference highlights two competing approaches: one proactive and integrated by default, and the other more manual and on-demand.

The launch comes amid intense competition between the two entities. OpenAI recently unveiled GPT-5, while Anthropic, having upgraded its Opus model to version 4.1, is reportedly seeking to close a funding round that would value the company at nearly $170 billion. Memory functions are part of this battle to retain users, considered a key lever to increase engagement and recurring use.

While the implementation is still partial, this development signifies a functional convergence that could quickly intensify, keeping pace with the announcements made by the two leading players in the conversational AI sector.