Anthropic now allows Claude to remember previous conversations to provide better context.
Anthropic’s Claude gets a memory function and can remember previous conversations. This way, the AI chatbot can remember someone’s preferences, the context of the project they’re working on, and their main priorities. This should lead to better responses, and the feature is “completely optional”.
Customized Memory for Work
According to the announcement, the feature seems to focus primarily on work-related tasks. Anthropic states that Claude “remembers project details, client information, and team processes.” A separate memory module is created for each project, ensuring that product launches and client files, for example, remain separate from each other.
Users get an overview of all stored memories, which they can view and adjust themselves. You can also tell Claude what to remember and what not to. This sounds similar to the memory function in other popular AI models like ChatGPT.
Incognito as an Alternative
In addition to memory, Anthropic is also introducing an incognito mode. Chats in this mode do not appear in the conversation history and are not added to the memory. This is mainly intended for confidential brainstorming sessions or conversations that are separate from ongoing projects.
The rollout is being implemented in phases to ensure safety and reliability. Currently, only Team and Enterprise subscribers can use the memory function. Incognito chatting is available for all users.