TL;DR: OpenAI’s new ChatGPT ad format, which shows a sponsored message while an image you requested is being generated, appears to be live now after being announced about a month ago. It’s not tied to what you asked ChatGPT to draw, which is the detail advertisers and users should both be paying attention to.
If you’ve asked ChatGPT to generate an image lately and noticed an ad pop up while you waited, you weren’t imagining it.Ad tech researcher Juozas Kaziukėnas spotted the format live in his own ChatGPT conversation on Monday and posted about it on X, noting that OpenAI had previewed this exact feature about a month earlier. The ad carries a simple, almost disarmingly honest headline: “Here’s an ad while we create your image.” Once the image finishes rendering, the ad disappears, like it was never there.
What makes the sighting notable isn’t the format itself, which is a fairly obvious thing to build once you accept that ChatGPT now has dead air to fill. It’s what Kaziukėnas pointed out about targeting. His prompt had nothing to do with dogs or pet food, and yet the ad he saw wasn’t tied to the subject of his image request at all. For a company that has spent years building a reputation on understanding exactly what you mean, serving an ad with no apparent connection to the conversation reads less like a targeting feature and more like a placeholder waiting to be filled in with something smarter later.
That’s speculation on my part, to be clear, because OpenAI hasn’t put out an official statement confirming the rollout, its scope, or how targeting actually works here. The company didn’t respond when reporters asked. What we do know is that this isn’t OpenAI’s first swing at advertising inside ChatGPT. It’s been layering in ad formats for months: wider display ads, multi-image ad units, a conversion tracking page, and reportedly sponsored agents that can act on a brand’s behalf inside a chat. Kaziukėnas, for what it’s worth, has a track record of spotting these rollouts early, having also found the new ChatGPT business hub days before it was formally announced.
The ad he saw simply read “Here’s an ad while we create your image,” with no connection to his prompt’s subject matter, per his post describing the sighting.
Here’s why the placement itself is clever, whatever you think of the ethics. Image generation is one of the few moments in a ChatGPT session where the user is genuinely stuck waiting, staring at a progress indicator with nothing to do. Every other ad-supported product on the internet has spent two decades fighting over exactly that kind of dead time, whether it’s a loading screen, a video pre-roll, or the three seconds before a page finishes rendering. OpenAI finding its own version of that gap was probably inevitable the moment it decided advertising belonged in the product at all.The open questions are the ones that actually matter for advertisers evaluating whether to spend money here. We don’t know pricing. We don’t know which countries or account tiers see these ads. We don’t know if free-tier users get more of them than paying subscribers, or whether this is a limited test that could vanish next week.
For the SEO and search marketing crowd, this is worth watching for a reason that goes beyond ChatGPT specifically. Every major AI product that answers questions directly is eventually going to need a business model that doesn’t rely purely on subscriptions, and advertising inside the answer, or inside the wait for the answer, is the most obvious lever available. Google has been circling the same problem for years with its own AI Overviews, where letting people opt out hasn’t stopped the click-through debate, and the pattern across both companies is the same: AI answers are getting monetized, it’s just a question of how visible that monetization ends up being to the person asking the question.
Marketers who’ve spent years optimizing for where an ad sits on a search results page are going to need a new mental model entirely once the ad sits inside a generative wait state instead of next to a list of links. There’s no click-through rate benchmark for an ad that disappears the moment your image finishes loading.
There’s a revenue pressure story underneath this that’s easy to miss if you only look at the feature itself. Running inference for every single ChatGPT conversation costs real money, and image generation is among the most compute-intensive things the product does. A subscription model alone has limits on how much it can subsidize that cost at ChatGPT’s current scale, which is exactly the math that pushed display advertising into search results, free email, and basically every other free or cheap internet product that came before it. OpenAI testing ads specifically during the most expensive moment in a session, the generation wait itself, isn’t a coincidence. It’s the company finding the exact spot where the cost of serving you is highest and deciding that’s also the moment worth monetizing first.
Compare that to how Google handled the equivalent problem with traditional search. It took Google years to settle into the ad formats and disclosure norms that regulators and users eventually tolerated, and it’s still fighting antitrust battles over exactly how those ads get placed and priced. OpenAI skipping straight to ads inside a generative wait state, barely two years after ChatGPT became a mainstream product, suggests the company isn’t interested in replaying that slow, decade-long evolution. It’s trying formats fast, watching what sticks, and presumably figuring out the targeting and disclosure rules as it goes rather than before it ships.
The bottom line
Bottom Line: This is a small, easy-to-miss feature today, but it’s a preview of where AI chat interfaces are heading commercially. Watch for OpenAI to say more about targeting and pricing in the coming weeks, because right now it looks like a test that forgot to turn on the part that makes it relevant to what you actually asked for.
The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.
