Matthew Forzan, Co-Founder and Managing Director of Yoghurt Digital.
For months, people have been waiting for ads to arrive on AI platforms. Not because they want them, but because they expect them. Advertising has long been the price of free digital services, and users have grown accustomed to seeing sponsored links on Google, social media, and various search engines. What’s different is that, until now, ChatGPT has remained conspicuously ad-free – and that absence has grown trust among users.
OpenAI’s move to test advertising inside ChatGPT represents more than a commercial pivot. It introduces a psychological shift. Users may tolerate ads, but they are far less forgiving when those ads appear to influence the answers themselves. The fear is not interruption, but bias.
Matthew Forzan, Co-Founder and Managing Director of Yoghurt Digital, believes that concern could drive some users away. “People expect ads on Google, Facebook, and Instagram, but they’ve never expected them in ChatGPT,” he says. “Once advertising enters the conversation, users will start questioning why they’re seeing certain answers.”
The question of objectivity sits at the heart of the debate. Does advertising automatically compromise neutrality, or does it simply change the way that neutrality is perceived?
Forzan is careful not to assume bad faith. “ChatGPT has an explicit answer independence policy, which states that ‘Ads do not influence the answers ChatGPT gives you’.” he says. “But we have to see how these ads work in practice.” He offers a simple example: if a user asks ChatGPT for hotel recommendations while planning a trip, receives a balanced, unbiased response, but is then shown a hotel ad immediately afterwards, was the answer still objective?”
Technically, yes. But psychologically, it becomes blurrier.
That distinction between front-end experience and back-end mechanics is where the real challenge lies. From a marketer’s perspective, ads cannot exist without prediction. “There has to be a model behind it,” Forzan explains. “You put in X dollars, you expect Y impressions, Z level of engagement. That’s how digital advertising works.”
Google has spent more than two decades building and optimising that infrastructure. Its ad ecosystem is built on intent signals, attribution models, and measurement frameworks that allow brands to justify spend, even when performance isn’t perfect. OpenAI, by contrast, is entering the market without those historical behaviour datasets or established ROI benchmarks.
This creates tension on both sides. Users worry that ads will subtly shape answers. Marketers worry that, without robust analytics, they won’t be able to justify investment. OpenAI must somehow reassure both, proving that AI responses remain untouched by commercial influence, while also offering advertisers enough predictability to open their wallets.
At the same time, this shift reflects a broader change underway in digital marketing. The lines between search, retail media, and affiliate marketing are rapidly blurring. As AI-driven discovery models grow, marketing spend is being reshuffled to follow attention earlier in the decision-making process – before consumers even realise that they are in “search mode.”
In that sense, ChatGPT is not simply adding ads; it is redefining where influence begins. But influence without trust is worthless. If users feel that answers are shaped by who paid the most, even subtly, the platform risks undermining the very value that made it powerful in the first place.
Forzan believes that trust between platforms and brands are what sets successful brands apart from the rest. “People are tired of hearing about another data leak or platforms using their personal information for profit,” he explains. “In a world where personal data often feels beyond individual control, trust between users and the platforms they rely on has become essential.”
The success of AI advertising will hinge on transparency, not just about what is sponsored, but about how objectivity is protected behind the scenes. If OpenAI can clearly separate prediction models for ads from reasoning models for answers, it may convince users to stay and marketers to spend.
If it can’t, ads won’t just monetise ChatGPT – they may push people off the platform altogether.
As new AI tools race to win mainstream adoption, their business models are becoming just as important as their capabilities. Google’s AI mode, for now, is relatively ad-free – but few doubt that advertising will arrive, as it already has in other markets. Users are accustomed to this trade-off with Google; ads are the price of entry, not a dealbreaker.
By contrast, Anthropic’s Claude has taken a more pointed stance, recently rolling out cheeky ads that promise an ad-free experience while subtly suggesting that this will remain the case. The catch, of course, is where the cost shows up instead: a capped free tier and a premium price tag for more advanced access. Together, these approaches reveal a familiar tension in tech, whether users pay with their attention or their wallets, and a growing expectation that “free” AI is never truly free.
This experiment fast approaches. Whether it becomes the future of digital advertising or a cautionary tale will depend on how well OpenAI balances revenue with the one asset no AI can regenerate once lost: trust.
