In a move that will shock absolutely nobody who has spent more than five minutes thinking about how free software works, OpenAI has quietly updated its privacy policy to enable marketing cookies by default for free ChatGPT users. The goal, per a report by Wired, is pretty straightforward - use that tracking data to nudge free users toward pulling out their credit cards.
What's actually happening here
When you use ChatGPT without paying, OpenAI now drops marketing cookies on your browser by default. These aren't the cozy, chocolate-chip kind. These are the surveillance-capitalism kind, designed to build a profile of your behaviour so the company can serve you targeted messaging aimed at converting you into a paying subscriber.
You can opt out, because regulators in various parts of the world would very much like companies to give you that option. But the key word is default. Most people don't dig through cookie settings. That's not a conspiracy theory, it's just human nature, and every growth team at every tech company knows it.
Why this matters beyond the obvious "big tech bad" angle
Here's the thing - ChatGPT's free tier has always been a funnel. That was never a secret. OpenAI is a company burning through cash at a truly spectacular rate, and converting free users to paid subscribers is literally how it survives. Fair enough.
But there's something a little rich about an AI assistant that you talk to like a trusted confidant also quietly building an ad-targeting profile on the side. You're asking it about your health, your relationships, your job anxieties - and the infrastructure around it is going, "cool, cool, now let's figure out how to sell this person a Plus subscription."
It's not sinister, exactly. It's just very, very normal Silicon Valley, which might actually be the more unsettling thing.
What you should do about it
If this bothers you, go into your ChatGPT settings and opt out of marketing cookies. It takes about 90 seconds and will make you feel briefly powerful in a world that rarely offers such opportunities.
If it doesn't bother you, that's also completely valid. Just go in with eyes open knowing that the free lunch comes with a side of behavioural tracking.
OpenAI isn't doing anything technically out of the ordinary here. It's doing exactly what basically every other free-tier tech platform does. The only mildly scandalous part is that we somehow still expect AI companies to be different - and they keep very efficiently proving us wrong.
