“The most profitable ad in Pepsi history.” The most voted comment in YouTube of the ad generated with AI by Coca-cola for Christmas 2025 suggests something: a popular rejection of advertising made with AI. Is this true? A new study from the University of Zaragoza on the effect of artificial intelligence on advertising points in that direction.
The researchers’ conclusion is that customers avoid services advertised with AI-generated images, especially in companies that offer pleasurable experiences—such as hotel vacations—or that force high-involvement decisions. The reason? Artificially generated images are interpreted as unreliable.
Its four authors explain to Xataka that “consumers value real images more because they show a faithful image of the product or service and they distrust companies that use images created with AI because they seem less professional or hide reality.” However, recent studies show that images created with AI can be equally effectiveand easier for companies to obtain, especially when consumers do not know that they are not real, they clarify.
What AI gives you, AI takes away
Using AI in an advertisement conveys a feeling that “the brand makes little effort, especially in luxury and beauty brands,” explains Lucía Caro Castaño, professor at the Department of Marketing and Communication at the University of Cádiz. After the Christmas controversy, Coca-Cola was forced to share how did you make the announcement to show “all the effort and investment it had required in terms of people.”
Caro points to savings in personnel as one of the reasons why content made with AI generates disgust. Coca-Cola has recognized The Wall Street Journal that producing your typical Christmas advertisement has gone from needing a year to a month, recognizing savings in costs and time. However, the creation of spot forced to enormous human work to fine-tune AI-generated images.
Coca-Cola is not the only company that has discovered the advertising limitations of AI. Dell share your experience: “We’re very focused on getting the most out of a device’s AI capabilities, but what we’ve learned this year, especially from a consumer perspective, is that they don’t buy based on AI. In fact, I think AI probably confuses them more than it helps them understand a specific outcome,” argued a few months ago Kevin Terwilliger, Chief Product Officer at Dell.
There are several reasons for this rejection of advertising AI: the feeling of “already seen”, which penalizes the lack of originality and creative effort; and the perception of “dehumanization” transmitted by excessively robotic content, explains Patricia Coll, doctor in Communication and professor at EAE Business School.
Diana Gavilán, professor of Marketing at the Complutense University of Madrid, highlights the benefits of AI in automatable tasks in advertising and digital marketing: “The problematic thing is when it replaces a human. If a robot serves me but you want to convince me that it is like a human, there is a drop in confidence.”
According to researchers at the University of Zaragoza, their study shows that real images are particularly effective when it comes to a product or service with high involvementthat is, the consumer wants faithful images when the decision they make is important. Real images are also better than those generated with AI to publicize hedonistic products or services because they allow “a better assessment of what the personal experience will be like.”
On the other hand, when the products are utilitarian and low-involvement, images generated with AI are effective.
In some sectors it is advisable to use commercial images made with AI, such as schools and social entities to avoid showing real children to protect their privacy, scientists highlight. The professor of Marketing at the University of Alicante, Ana Belén Casado, adds that not all consumers or all brands reject AI: “It depends a lot on the type of product, good, service or idea that is being marketed and the differential value proposition of each brand.”
For Gavilán, AI is like the Thermomix: a tool with which you don’t do everything in the kitchen, “but you can use it and it is at your disposal, depending on how you use it, it will be better or worse.” In his opinion, the Coca-Cola ad was “a strategic mistake” for wanting to make the same old ad with AI instead of making a different story with that technology.
Brands taking a step back with AI?
Before Coca-Cola, the clothing brand H&M had already launched a campaign with real models and “digital twins” generated with AI.
Although all images generated by AI are labeled so as not to confuse them with real ones and the models have image rights Regarding his digital copy, Caro highlights that “we do not know exactly what this contract has been like in terms of the rights to his own image that exceed those models, nor will it affect photographers and the rest of the workers who make these campaigns possible.”
This innovative campaign was quite small, around a line of denim clothing, and the head of AI at the Swedish multinational, Linda Leopoldleft the company shortly after the campaign. “We don’t know where H&M will continue next, especially with all the controversy generated,” says Caro. Gavilán’s vision is that AI will continue to be implemented and that it will be applied more in areas “where it is very relevant.”
Despite his water and energy consumptionthe environmental NGO WWF in Denmark launched a campaign titled “The hidden cost” in April 2025 to denounce the environmental impact of eleven different products. It was made entirely with AI. In Spain the first advertising agency focused on AI, AI::gencyhas worked with brands such as Nissan, Seat, Cushla and Ebro.
Other brands have chosen publicly reject the AI. In his campaignWhy don’t we get on the AI bandwagon?” in February 2024, the browser Vivaldi announced that it would not be incorporating AI “for the time being.” The reasons given by the company were copyright and privacy violations, as well as “plausible-sounding lies” generated by AI.
At the advertising level, doveUnilever’s personal care brand, has committed to “under no circumstances use images generated by artificial intelligence to represent or replace real people.” In 2006, Dove, which advertises itself as “real beauty,” launched its spot ‘Evolution’ in which it denounced the use of Photoshop in beauty models. Since then Dove avoids its use for “distort” reality.
The conclusion of the researchers from the University of Zaragoza is that “we must learn to develop our human qualities that differentiate us from AI and that provide a human valueemotional and social that technology cannot yet achieve, such as empathy. “AI is capable of performing analytical and creative tasks that until now only people could do, so its integration and regulation represents a challenge for society and institutions.”
“Perhaps in the first phase, using AI served to distinguish itself as innovation, but AI is no longer perceived as cutting-edge and we are tired of it,” highlights Lucía Caro. As technology analyst Neil Postman, whom Caro quoted in our interview, said, “at this point, not being aware that any technology is accompanied by a program of social change, maintaining that technology is neutral, assuming that it is always a friend of culture, is pure and simple stupidity.” At this point it was 1985.
In Xataka | Three AIs clashed in ‘War Games’. 95% of them resorted to nuclear weapons and none ever surrendered
GIPHY App Key not set. Please check settings