If you’ve opened ChatGPT recently and noticed something like “Testing ads in ChatGPT”, it probably made you pause. Not because anyone loves adverts, but because it raises a very specific worry. As a paid media agency at Honcho, we spend a lot of time thinking about how advertising shapes platforms, and how platforms shape behaviour, so this one is worth paying attention to.
Is ChatGPT about to start “recommending” things because someone paid for it?
Thankfully, what’s being tested, at least in the versions people are seeing, looks more like an ad slot in the interface than ads baked into the conversation. That distinction matters more than it sounds.
The quick version
ChatGPT is testing sponsored ads for some users.
They do not show up inside the AI’s reply. Instead, they appear separately on the screen, usually underneath the response, and labelled as sponsored.
So you read the answer first, and if an ad appears, it’s basically a separate block after the fact.
Where the ads show up
If you’re in the test group, the flow is roughly:
-
you ask a question
-
ChatGPT gives you a normal answer
-
then you may see a “Sponsored” card below it
It’s the difference between “here’s my advice” and “by the way, here’s an ad”. Still not charming, but far less sneaky.
Does this affect recommendations?
This is the bit everyone cares about.
The idea, as described, is that the AI response is generated normally, and the ad is a placement the platform shows separately. So if you asked “recommend a pen for office use”, ChatGPT might suggest a few sensible options, and then you might see an advert for a retailer underneath.
That means you can ignore the ad and still get a useful answer, in theory.
In practice, people will judge this on behaviour, not promises. If the assistant starts sounding like it’s trying to sell you something, it won’t matter how neat the “Sponsored” label looks.
Live Q&A: I asked ChatGPT directly
I wanted to see how ChatGPT itself explains the update, and what it can and can’t do about it, so here’s a Q&A I’ll keep adding to as I go.
Question 1
Why has ChatGPT started testing ads?
ChatGPT:
Because AI is expensive to run, and the free tier has to be funded somehow. Ads are a way to generate revenue without putting more features behind a subscription.
Question 2
What are the advantages of introducing ads?
ChatGPT:
It helps fund the free version of ChatGPT, keeps access open to more users, and creates new opportunities for brands as AI becomes part of how people discover products.
Question 3
What are the downsides or risks?
ChatGPT:
The main risk is trust. If ads ever start influencing answers, or feel intrusive, users will notice quickly. There’s also a wider concern about too much power sitting with a small number of AI platforms.
Question 4
What would advertising look like in an agentic commerce world?
ChatGPT:
Instead of bidding on keywords, brands would optimise product data, pricing, and availability so AI agents can recommend them. Visibility would depend less on ad copy, and more on structured feeds, trust signals, and real performance.
Question 5
It will probably be mixed. If ads stay clearly labelled and separate from answers, most users will tolerate them. If they start shaping responses, the backlash will be quick.
Who’s likely to see ads
This looks like a test rather than a universal rollout, which usually means some people will see it and others won’t, even if they’re using the same app.
Broadly speaking:
-
free users are the most likely to see it
-
paid plans tend to stay ad-free
-
it may vary by region, device, and account
If you’re not seeing ads, it doesn’t necessarily mean they’re not happening, just that you’re not in the group being tested.
Why they’re doing it
AI is expensive to run. That’s not marketing spin, it’s just true. And ChatGPT has a massive number of free users.
So ads are one way to keep a free tier alive without pushing everything behind a subscription. It’s basically the same logic as most of the internet, for better and worse.
From an advertiser’s point of view, it also hints at what could be coming next. If AI platforms become recommendation engines as well as search tools, understanding how paid visibility works inside them will matter, which is exactly the sort of shift we think about in PPC strategy and services.
A bit of scepticism, because it’s healthy
Even if ads are cleanly separated today, it’s worth being cautious about where this goes.
Once advertising becomes part of the business model, the pressure tends to creep. First it’s “Sponsored” cards under answers. Then it becomes more prominent. Then it becomes harder to tell what’s genuinely recommended versus what’s commercially convenient.
There’s also a bigger market concern here. If one or two AI platforms end up dominating search, shopping, and “what should I buy” decisions, it concentrates an enormous amount of influence in very few places. Ads can accelerate that, because the biggest players have the data, the distribution, and the budget to keep users inside their ecosystem. Over time, that can make it harder for smaller tools, publishers, and independent retailers to compete.
None of that is guaranteed, but it’s the direction people will be watching.
The bottom line
If you’re on a paid plan, you probably won’t notice any change.
If you’re on the free tier, you might start seeing sponsored blocks under some answers.
Either way, the thing to keep your eye on is simple:
-
Are ads clearly labelled?
-
Do they stay separate from the actual response?
-
Does the assistant still feel genuinely helpful, even if you ignore every ad?
If the answer stays useful without the ad, most people will shrug and move on. If it starts feeling like the ad is steering the conversation, that’s when trust starts to go.