ChatGPT's memory feature now automatically learns about you from every conversation, building an evolving profile that influences how the AI responds to your questions. Unlike the original memory system that required you to explicitly ask the chatbot to remember things, the updated version uses what OpenAI calls "dreaming" to continuously refine its understanding of your interests, preferences, and personal details without your direct input.

How Has ChatGPT's Memory Changed?

When ChatGPT first introduced memory in April 2024, the system was rigid and manual. You had to explicitly tell the chatbot "remember this" about something, and those memories stayed frozen in time. If you mentioned an upcoming party, ChatGPT would forever assume it was still in the future, even after the date had passed, until you manually corrected it.

An update in June changed everything. ChatGPT now uses flexible, self-updating memories that evolve over time. The system automatically decides what's important enough to remember from your conversations, without requiring you to make manual requests. This means the chatbot continuously refines what it thinks it knows about you every time you interact with it.

OpenAI designed this system with three main goals in mind. The company stated that the three main aims of memory are to carry forward useful context, to follow preferences and constraints, and to stay current over time. In theory, this should deliver responses that feel more personalized and helpful, but the tradeoff is that you have less direct control over what gets remembered.

What Kinds of Details Is ChatGPT Remembering About You?

ChatGPT's automatic memory can capture a surprisingly broad range of information. If you spend a week researching how to replace a light switch, the AI might decide you're a DIY enthusiast and start framing answers around home improvement projects. If you look up restaurants and attractions around a specific location while planning a vacation, ChatGPT might incorrectly conclude that you actually live there and adjust its recommendations accordingly.

The memory system stores information about your interests, your preferences for how you like answers delivered, and other personal details that emerge from your conversations. All of ChatGPT's responses now come with this memory context baked in, meaning the AI is constantly filtering its answers through its assumptions about who you are.

How to View and Control What ChatGPT Remembers

You have several options for managing ChatGPT's memory, ranging from fine-tuning what gets stored to wiping everything clean. Here's how to take control of your AI profile:

• Access Your Memory Summary: On the web app, click your profile icon in the bottom left, select Personalization, and scroll down to the Memory section. On mobile, tap the menu button at the top left, then your profile icon, then Memory. You'll see a summary of what ChatGPT thinks it knows about you.
• Make Corrections: In the Memory section, you can highlight any information and choose "Make a correction" to fix inaccuracies, or select "Don't mention this again" to prevent the AI from using that detail in future responses.
• Issue Direct Commands: Use the prompt box at the bottom of the Memory section to issue specific commands like "forget that I'm a journalist" or any other detail you want removed from the AI's understanding of you.
• Delete All Memories: To wipe your entire memory profile, select the three dots at the top of the Memory section, choose Delete, and turn off memory. This starts you fresh with ChatGPT, though your past conversations remain in your chat history and will be re-indexed if you turn memory back on.
• Disable Memory Entirely: Toggle off the memory feature to prevent ChatGPT from remembering anything new. Note that this doesn't erase existing memories; it just prevents the AI from referencing them or adding new ones.
• Use Temporary Chats: For conversations you don't want stored in memory at all, use ChatGPT's temporary chat feature. On the web, click the dotted speech bubble icon at the top right to start a temporary chat. Nothing in these chats gets committed to memory, and they're forgotten as soon as you close them.

The simplest approach, of course, is to avoid telling ChatGPT anything you don't want remembered in the first place. But for most users, the memory management tools offer a practical middle ground between convenience and privacy.

Why This Matters for Your AI Experience

The shift from manual to automatic memory represents a fundamental change in how AI chatbots interact with users. While the convenience of not having to explicitly request memory is appealing, it also means the AI is making judgments about your identity and preferences without your explicit approval. ChatGPT might misinterpret your research habits, make incorrect assumptions about your location or profession, or build a profile that doesn't actually reflect who you are.

The stakes are higher now because these memories directly shape the answers you receive. An AI that thinks you're a DIY expert will frame technical questions differently than one that knows you're a beginner. An AI that incorrectly believes you live in a city you were just visiting will give you location-specific recommendations that don't apply to your actual life. Understanding what ChatGPT remembers about you is essential to getting accurate, relevant responses.