OpenAI supercharges ChatGPT memory in major leap toward AGI — Now remembers everything, and feels almost human

ChatGPT just got a major memory upgrade—and it might be the biggest step yet toward something that actually feels like a smart assistant, not just a chatbot. OpenAI announced today that ChatGPT can now reference your entire chat history to respond more naturally, understand your preferences, and actually remember what matters to you.

Whether you’re writing, learning, or just looking for advice, it now brings more context into the conversation—like talking to someone who actually listens. For many, this feels less like a feature update and more like a leap toward artificial general intelligence (AGI).

OpenAI confirmed the update on X, posting:

“Starting today, memory in ChatGPT can now reference all of your past chats to provide more personalized responses, drawing on your preferences and interests to make it even more helpful for writing, getting advice, learning, and beyond.”

Starting today, memory in ChatGPT can now reference all of your past chats to provide more personalized responses, drawing on your preferences and interests to make it even more helpful for writing, getting advice, learning, and beyond. pic.twitter.com/s9BrWl94iY

— OpenAI (@OpenAI) April 10, 2025

Until now, ChatGPT could only recall limited user-provided details—things like your name, writing style, or favorite topics—stored through its existing memory feature. This update pushes things further. It doesn’t just remember what you told it to. It remembers everything you’ve said in prior chats and can apply that context automatically. You’re no longer starting from scratch each time.

ChatGPT Can Now Remember Everything—And It Feels Almost Human

The update gives ChatGPT a memory that spans your entire chat history—not just isolated facts. ChatGPT now remembers more than just facts—it remembers you. By tapping into your entire conversation history, it responds in ways that feel familiar, natural, and surprisingly human.

A Chatbot That Actually Feels Like It Knows You

This change means ChatGPT can now build on earlier conversations—without being asked. Say you asked for writing advice in February and brought up a different topic in March. If you return in April, ChatGPT could still remember your tone preferences or learning goals from those earlier chats. As OpenAI puts it, “New conversations naturally build upon what it already knows about you, making interactions feel smoother and uniquely tailored to you.”

People are already noticing. One user, @kimmonismus, shared this on March 28:

“Improved memory for ChatGPT is coming. It now remembers more from the last conversations it had with you. Memory is such an underrated feature. I love it!”

That was before the feature even rolled out officially—proof that users have been waiting for ChatGPT to feel less forgetful and more like a real conversation partner.

Improved ChatGPT Memory: Users Still Hold the Keys

OpenAI seems aware that the word memory can raise eyebrows, especially when privacy is on the line. Their post includes this reassurance:

“As always, you’re in control of ChatGPT’s memory. You can opt out of referencing past chats, or memory altogether, at any time in settings.”

If you’ve already turned off memory, ChatGPT won’t start pulling in old chats unless you change your settings. You can even ask it what it remembers—or delete specific items. And for one-off conversations where you don’t want any context remembered or used, “temporary chat” is still an option.

That level of transparency is crucial, especially as AI assistants start acting more like long-term collaborators than short-term tools.

Who Gets It and When?

The memory upgrade starts rolling out April 10 to ChatGPT Plus and Pro users—except in regions like the EEA, UK, Switzerland, Norway, Iceland, and Liechtenstein, where data regulations require a bit more paperwork. OpenAI says Team, Enterprise, and Edu accounts will get access in the coming weeks. You’ll know it’s available once a notification appears in ChatGPT.

A Bigger Shift in AI Behavior

This update doesn’t just make ChatGPT more convenient. It signals a shift in how AI is being positioned—less like a question-answering machine, more like a personal assistant that gets smarter the more you talk to it. The experience starts to feel continuous, not fragmented. That’s a big deal for anyone who relies on ChatGPT to brainstorm ideas, study, plan, or work on long-term projects.

Of course, not everyone’s thrilled. Some X users have raised questions about how much ChatGPT really remembers. A post from @AISafetyMemes back in December pointed out:

“Deleting a past conversation does NOT erase memories remembered from that chat.”

That lines up with what OpenAI has said: memories are stored independently of specific conversations. If you want something gone, you’ll need to delete it from memory directly.

Still, the rollout is a bold move. As other AI players like Google’s Gemini and Anthropic’s Claude try to win over users with smarter, more context-aware bots, OpenAI’s memory revamp raises the bar. The real question is whether users will embrace this new version of ChatGPT—or get spooked by the idea of a chatbot that actually remembers everything.

This move fits into a larger pattern. On March 31, OpenAI shared that it plans to release a new open-weight language model with advanced reasoning capabilities in the coming months. While details are still sparse, the announcement suggests that OpenAI is actively broadening its approach—working on both smarter personal assistants like ChatGPT and more transparent, open models that the wider community can build on.

Together, these developments point to something bigger than just feature updates. OpenAI is steadily shaping its products around long-term goals—context, reasoning, and maybe one day, general intelligence.

For now, ChatGPT Plus and Pro users (outside restricted areas) can give the new memory a spin. Whether it’s helpful or a little too helpful—that’s up to you.