If you opened the ChatGPT app this week and noticed it acting a little more familiar, you're not imagining things.
OpenAI pushed an update to its mobile app that expands the "memory" feature first tested on a limited basis, and the rollout is happening with almost no fanfare.
Just a quiet toggle that changes how the most popular AI app on your phone interacts with you.
ChatGPT can now carry context across separate conversations on mobile, not just within a single thread.
That means it might recall that you're training for a half marathon, that your kid is allergic to peanuts, or that you prefer your code in Python rather than JavaScript.
Previously, each chat started closer to a blank slate unless you manually fed it background.
Now the app builds a running profile that persists.
The feature lives in your settings, and OpenAI says you can view, edit, or delete what it remembers at any time.
That last part matters more than most users realize.
A memory system is only as trustworthy as its controls, and the company is clearly trying to get ahead of the privacy backlash that's been brewing since the feature first appeared.
Why should you care if you're not a power user?
Because this is the moment ChatGPT stops being a search box and starts being an assistant.
Your phone already knows your calendar, your location, and your music taste.
The AI app on your home screen is now competing for the same cognitive real estate, and it's doing it with a memory layer that Google and Apple have been slow to match.
Apple's Siri overhaul and Google's Gemini push are both betting that context is the next battleground.
OpenAI is effectively saying: we already have the users, and now we're giving the app a brain that doesn't reset every time you close it.
Whether that's convenient or unsettling depends entirely on how much you trust a chatbot with your personal details.
Ask for dinner ideas on Monday, and by Friday the app might remember you're vegetarian and that you hate mushrooms.
Request a workout plan and it can factor in the knee injury you mentioned two weeks ago.
That continuity is the difference between a tool you use and a tool that knows you.
They can latch onto outdated information, or worse, misremember something you said in passing.
If the app starts acting on a stale detail, you'll need to dig into settings and prune it manually.
OpenAI hasn't solved the "garbage in, garbage out" problem, it's just made it stickier.
For most American users, the update will roll out gradually over the coming weeks.
If you don't see it yet, check your app version and give it time.
The feature is opt-in in some regions and on by default in others, so it's worth a quick look at your settings regardless of where you stand on the convenience-versus-privacy tradeoff.
My take: this is the update that quietly moves ChatGPT from novelty to utility, and it deserves more scrutiny than it's getting.
Convenience always wins in consumer tech, but memory without transparency is how trust gets spent.
Final Thoughts
Check what your app remembers, and decide for yourself whether the trade is worth it.