OpenAI’s move to show ads in its chatbot marks a turning point for how AI makes money and for the balance between commercialization and trust. With insight from IAB’s VP of AI, Caroline Giegerich, we look at what it means for users and advertisers.
What ads on ChatGPT might look like (ChatGPT)
For all the talk of AI as the future of everything, one question has loomed awkwardly in the background: how does any of this actually make money at scale? Now we have an answer. Or at least the beginning of one.
OpenAI has confirmed it will begin showing advertising to users of the free version of ChatGPT, placing sponsored messages at the bottom of the interface and targeting them based on what users are asking – not what the model itself says. Paid tiers will remain ad-free, while a cheaper subscription will introduce a hybrid model with ads plus extra features.
This isn’t just a product update. It’s a watershed moment for advertising and a stress test for whether conversational AI can be monetized without destroying the very thing that made it useful in the first place.
As Caroline Giegerich, VP of AI at IAB, puts it: “Platform monetization meets user experience and it’s a delicate balance.”
Advertisement
LLM platforms don’t have much choice. Running models at a global scale costs billions and investors aren’t funding AI out of goodwill. Advertising is the most obvious lever. But unlike search or social, chatbots live and die by flow. Interrupt that and the user leaves.
“If ads feel interruptive rather than native to the conversational experience, users will abandon the platform,” says Giegerich. “In a market where everyone’s racing for dominance, losing stickiness means losing the game.”
That’s the central tension. A banner ad dropped into a conversation about grief, health or education would feel grotesque, but a contextual offer – a rug discount when you’re literally asking how to furnish a living room – could feel helpful rather than hostile.
“‘Ad’ doesn’t have to mean ‘bad’ if the recommendation fits naturally into what you’re already asking about,” says Giegerich. This is why OpenAI’s promise to exclude sensitive categories, avoid ads for children and keep sponsored content separate from model responses matters more than any revenue forecast.
If there’s one rule that cannot be bent in conversational AI advertising, it’s disclosure. “Transparency isn’t optional – it’s the foundation,” says Giegerich. “Consumers need to know when they’re seeing an ad. Period.”
That’s easier said than done in a medium where responses feel authoritative by default. A sponsored suggestion that looks like an answer risks eroding trust not just in ads, but in the system itself. This is why the industry is already scrambling to catch up.
“Existing advertising guidelines weren’t written for conversational AI. At IAB, we’re developing transparency frameworks specifically for this reason.”
Early platforms that label clearly, give users control and draw hard lines around what can and can’t be monetized will earn credibility. Those that blur the lines may win short-term revenue but lose long-term belief.
For brands, the bigger headache may be proving that any of this works. “We saw over the holidays just how many consumers are using LLMs for purchasing decisions,” says Giegerich. “But traditional attribution models don’t work here.”
If a chatbot suggests three products and one is sponsored, what’s success? A click? A purchase later that day? The fact that the conversation continued?
“This isn’t about buying real estate,” she says. “It’s about earning relevance in context.”
That’s a fundamentally different mindset for advertisers raised on impressions, clicks and last-touch attribution. Chatbot advertising isn’t a slot – it’s an intervention. And the wrong one will feel immediately wrong.
Advertisement
There will be backlash. There always is. But Giegerich believes the long-term outcome will look familiar.
“Yes, there will be initial resistance, especially when ads feel interruptive, but if ads are clearly disclosed, contextually relevant and genuinely helpful, consumers will acclimate – just like they have with every other advertising evolution.”
The difference is that AI assistants come with a higher bar. They’re framed as helpers, not feeds. Abuse that role and trust collapses quickly.
Advertisement
Google is already moving in the same direction with ads in AI-powered search. Others will follow. The idea that AI could remain neutral, free and ad-less forever was always a fantasy.
The real question isn’t whether AI will carry ads, it’s whether advertising can adapt to a medium that punishes irrelevance instantly.
Get this right and conversational AI becomes one of the most powerful intent signals marketers have ever seen. Get it wrong and users will simply move on to the next assistant that feels less compromised.
For OpenAI, this is about funding the future. For advertisers, it’s about relearning restraint. For users, it’s a reminder that even the smartest machines still need paying for.
And in AI, trust is the only currency that really matters.
OpenAI