As generative AI platforms scale at extraordinary speed, the business realities behind them are coming into sharper focus. Massive infrastructure costs, hundreds of millions of users, and intensifying competition are forcing companies to answer a critical question: how do you sustainably monetise AI without undermining user trust?
OpenAI’s reported plans to introduce advertising into ChatGPT signal a potential shift not just for the company, but for the broader digital advertising ecosystem. If conversational AI becomes ad-supported, how will that reshape the sales funnel? Where will chatbot advertising sit between search and social? And what new ethical considerations emerge when ads are inserted into deeply personal, context-rich conversations?
In an episode of This Week in Business, Stefano Puntoni, Professor of Marketing at the Wharton School and co-director of the Wharton Human-AI Research Program, joined host Dan Loney to examine what advertising inside AI platforms could mean for consumers, regulators, and the future of digital marketing. Puntoni explored why the move may be inevitable, how it could disrupt Google and Meta’s dominance, and why transparency and trust will be decisive as AI platforms evolve.
Below follows an edited transcript.
Dan Loney: Recently, OpenAI announced it’s going to test bringing advertisements to ChatGPT. The move appears driven in part by mounting operational costs. It’s being viewed as a way to recoup some of the heavy investment in the platform.
Obviously, this is a business decision, but it raises plenty of questions. What was your reaction?
Stefano Puntoni: In many ways, it was inevitable. ChatGPT has hundreds of millions of users worldwide, and maintaining that kind of scale is incredibly expensive. Right now, the free tier generates no revenue. That simply isn’t sustainable long term.
There are really two paths forward. One is to push more users toward paid subscriptions. The other (and the most common model on the internet) is advertising.
Given how many people use the free version, advertising was always a likely outcome. And once one major AI platform moves in this direction, others may follow.
Dan Loney: Reports suggest the company plans to focus advertising on adults. But whenever we talk about internet platforms, concerns arise about children, especially around data collection and targeting. Is that a valid concern here?
Stefano Puntoni: Yes, it is. The age-verification systems currently in place for ChatGPT are relatively basic. If advertising becomes central to the business model, we may see pressure to strengthen those systems.
What makes chatbots different from traditional platforms is the depth of interaction. These systems collect incredibly rich conversational data. From those conversations, it may be possible to infer quite a bit about the user (including age) even if that information isn’t explicitly provided.
Whether that capability is used responsibly and whether regulators require stricter safeguards will be an important issue to watch.
Dan Loney: Let’s talk about the advertising ecosystem more broadly. How might this impact digital advertising?
Stefano Puntoni: This is potentially very significant. Digital advertising today is largely dominated by two formats: search advertising and social media advertising. Those spaces are led by companies like Google and Meta.
To understand where chatbot advertising fits, it helps to think about the “sales funnel.”
At the top of the funnel, consumers may not yet have a clear buying intention. Social media advertising operates there; you’re browsing content, talking to friends, watching videos, and ads are inserted into that experience.
At the bottom of the funnel is search advertising. When you type keywords into a search engine, you are signalling intent. That’s highly valuable because advertisers know you may be close to making a purchase. ChatGPT advertising likely sits somewhere in the middle.
On the one hand, it’s conversational and broad, like social media. On the other hand, users enter prompts that often reveal very specific needs or intent, much like search queries.
That combination could make chatbot advertising uniquely powerful. It blends context, conversation, and declared intent in ways we haven’t really seen before.
Dan Loney: So, this wasn’t a big surprise?
Stefano Puntoni: Not at all. The real question isn’t whether advertising would come; it’s how it will be implemented. The design decisions will matter enormously.
If ads feel intrusive or distort the user experience, there could be backlash. But if they’re carefully integrated (contextual, transparent, and relevant) users may accept them as part of the trade-off for free access.
Dan Loney: What will you be watching most closely?
Stefano Puntoni: Two things. First, how clearly OpenAI separates conversational responses from sponsored content. Transparency will be critical to maintaining trust.
Second, how conversational data is used. Because these systems capture nuanced, personal, and contextual information, the ethical bar is higher than for traditional web advertising.
Chatbots aren’t just another screen for banner ads. They represent a fundamentally different interaction model. That means the advertising model layered on top of them may also need to be fundamentally different.