AI memory: AI memory is a feature that lets an AI assistant retain facts across separate conversations, such as your role, preferences, and ongoing projects, so you do not have to repeat them. It differs from the context window, which lasts only for one conversation.
By default, a language model forgets everything the moment a conversation ends. Memory features bolt on a notebook. As you chat, the assistant saves facts it judges useful ('works in commercial insurance,' 'prefers short bullet-point answers,' 'is preparing a Q3 board deck') and quietly includes them in later conversations. ChatGPT, Claude, and Gemini all offer some version of this as of mid-2026, with differences in what is saved, whether you can see and edit it, and whether it is on by default. Check the vendor's current documentation.
Memory takes several forms. Saved memories are discrete facts you can review and delete. Chat history reference lets the assistant search your past conversations. Project or workspace features let you pin documents and instructions that apply to a whole set of chats. Most of these work by retrieving relevant notes and inserting them into the context window when a conversation starts, which is why memory can feel inconsistent: what gets pulled in depends on what the assistant judges relevant.
Memory is a convenience and a privacy consideration at once. Whatever the assistant remembers about a client, a colleague, or a confidential project persists and can surface in an unrelated chat, including one you share or show on a screen. Review the memory list periodically, delete what should not be there, and follow your organization's policy on whether memory belongs on work accounts at all.
Example at work
A consultant tells her assistant once that she writes for mid-market manufacturing CFOs and dislikes jargon. Weeks later, a fresh chat asking for a proposal outline comes back already pitched at that audience. Before a client call where she plans to share her screen, she opens the memory settings and removes a note about a different client's pricing.
Why it matters
Memory removes the most tedious part of using AI assistants: re-explaining who you are and what you are working on. It also creates a new place where sensitive information can live and leak, so knowing how to inspect and clear it is now part of basic AI hygiene at work.
Related terms
- Context windowThe context window is the maximum amount of text, measured in tokens, that a language model can consider at one time. It includes your prompt, any pasted documents, the conversation so far, and the model's reply.
- AI assistantAn AI assistant is a software application, usually built on a large language model, that helps you with tasks through natural conversation: drafting, summarizing, answering questions, analyzing files, and increasingly taking actions in other apps.
- Data privacyData privacy, in the context of AI, is the set of rules and practices that govern what information you put into an AI system, who can see it, how long the vendor keeps it, and whether it is used to train future models.
- Retrieval-augmented generation (RAG)Retrieval-augmented generation (RAG) is a technique in which an AI system first searches a set of documents for relevant passages and then gives those passages to a language model to answer from. It lets an assistant use your own, current information.
- Custom GPTA custom GPT is a tailored version of ChatGPT that you configure with your own instructions, uploaded reference files, and optional connections to outside services, then reuse or share without rewriting the setup each time.