The AI Trust Gap
Why Brands Keep Getting AI-Generated Advertising Wrong
Author: IMB Editorial Team
IMB Journal – International Marketing Board
Volume 1 | Issue 6
June 2026
Why Brands Keep Getting AI-Generated Advertising Wrong
In December, McDonald’s Netherlands published a Christmas commercial built entirely with generative AI. Within days, the ad had been pulled from YouTube, comments disabled, and a statement issued acknowledging the misstep. It was not an isolated event. Coca-Cola had already run two consecutive AI-generated holiday campaigns to a mixed and increasingly vocal reception. Across the industry, a pattern that looked like a string of unlucky one-off missteps in 2024 had, by the middle of 2026, become something closer to a rule: audiences can tell, and telling changes how they feel about the brand.
The strange part is that none of these companies were being careless. Every one of them had a clear business reason to use AI in exactly the way they did.
The Business Case Was Never the Problem
Generative AI genuinely does what its advocates claim. It compresses a production timeline that once took months into weeks. It removes the cost of location shoots, casting, and post-production. It lets a global brand localize a campaign across dozens of markets without commissioning dozens of separate shoots. On a spreadsheet, the decision to use it looks obviously correct.
What the spreadsheet doesn’t capture is what happens the moment a viewer notices. Industry research from the IAB and Sonata Insights found a striking gap on this point: the large majority of ad executives believed younger audiences responded well to AI-generated advertising, while actual survey data from those same audiences told a very different story, with adoption sentiment less than half what executives assumed. That gap between what marketing leadership believes and what the audience actually feels is where most of this year’s AI advertising failures have originated.
Audiences Aren’t Rejecting AI. They’re Rejecting Being Fooled.
It’s tempting to read the backlash as blanket resistance to artificial intelligence, but the pattern doesn’t support that reading. The same audiences who criticized fully AI-generated holiday ads have shown no comparable resistance to AI-powered recommendation engines, customer service tools, or personalized offers, so long as those tools stay in the background. Survey data from earlier this year found that a majority of consumers now doubt the authenticity of content they encounter online, and that a large share reduce their engagement with a brand the moment they suspect a machine produced what they’re looking at.
The common thread isn’t the technology. It’s the sense of being deceived about who, or what, made something meant to feel human. A holiday ad is supposed to traffic in warmth, nostalgia, and craft. When viewers pick up on subtly wrong facial movements or oddly generic-feeling scenes and realize no one actually made the thing they’re watching, the reaction isn’t neutral. It reads as a small betrayal, and the criticism that follows tends to be personal in a way that criticism of a mediocre but human-made ad rarely is.
The Brands Reading This Correctly
A different group of brands has taken the opposite position and, so far, appears to be benefiting from it. Dove committed publicly to never replacing people with AI in its advertising. Aerie, Equinox, and Almond Breeze have all run campaigns this year that directly mock AI-generated content, positioning themselves as a deliberate contrast to the trend. iHeartMedia launched a campaign built around a guarantee that its content, personalities, and music are made by humans, citing internal research showing an overwhelming preference for human-made media even among listeners who use AI tools themselves in their own lives.
None of these companies has abandoned AI. They almost certainly use it extensively for targeting, media buying, and campaign optimization, the parts of marketing a customer never sees. What they’ve drawn a line around is the customer-facing surface: the ad itself, the voice, the face, the story. That distinction, between AI as infrastructure and AI as the visible creative product, is turning out to be the difference between brands that absorb a one-time trend and brands that quietly accumulate a trust problem with every subsequent campaign.
What This Means in Practice
The lesson from this year isn’t that generative AI is unusable in advertising. It’s that the efficiency case for using it and the trust case for using it visibly are two entirely separate questions, and treating them as one is what has produced nearly every high-profile misstep so far. A brand can use AI aggressively behind the scenes while remaining deliberately, visibly human in front of the customer, and right now, that combination looks like the actual competitive advantage, not a compromise on the way to full automation.
Part of a three-part series on AI and advertising trust. Next: a case study on the McDonald’s Netherlands AI Christmas ad, from concept to removal.