What AI Doesn’t Understand About Advertising

The brands we remember rarely succeeded because they shouted the loudest. They stayed because they understood that people do not buy what they notice; they move toward what is trustworthy and emotionally legible.

That is why some adverts remain with us years later, long after the product is no longer relevant. They were not visually perfect, but they understood an essential about recognition. A child running barefoot through a compound just before supper. The sound of a kettle in the background while someone folds clothes. A father arriving home late and still making time. A grandmother portioning food with the kind of generosity that made the moment larger than the meal itself. Singing the Niko na Safaricom song on our way to school truly connected us. These scenes stay because they are less like invention and more like memory.

Psychology has long shown that familiarity lowers resistance. We are more likely to trust what is known, and trust is what advertising is always trying to buy. A familiar face, a street that looks like ours, a conversation that sounds like something we have overheard at home, these things tell us, quietly, that this message belongs in our world. That sense of emotional legibility matters far more than most brands will admit.

This is also why nostalgia works so well. Not because people want to return to the past, but because memory is one of the fastest routes to emotional trust. When a brand taps into a school-song chorus, the smell of rain on a Nairobi evening, or the particular warmth of Sunday lunch at home, it is doing more than storytelling. It is borrowing from emotional memory. Research published in the Journal of Consumer Research and more recent digital marketing studies both point to the same thing: nostalgia, when used authentically, deepens emotional engagement and makes people more receptive to brands.

That same instinct for recognition is what gave rise to the now-familiar ecosystem of affordable, accessible image libraries and stock media platforms. They emerged because not every brand could afford to stage full campaign shoots, yet every brand still needed access to the kind of imagery that helped bridge the gap between message and emotion. At their best, these libraries solved a practical problem without sacrificing the one thing that matters most in communication: believability.

They offered something simple and important: real human scenes, natural light, familiar gestures, ordinary moments that felt inhabited rather than manufactured. A woman laughing without looking posed. A child being held in a way that suggested history rather than direction. A kitchen that looked used rather than styled. Even when the faces were strangers’, the emotion often felt close enough to ours that we accepted the message.

This is where locally sourced media matters even more.

Using local photographers, videographers, neighbourhoods, textures, and cultural references does more than improve representation. It tells people the brand understands where they are, how they live, and what details matter to them.

In African markets, especially, local media is often the difference between being seen and being believed. Campaigns that use local references tend to be more trustworthy because they speak in a cultural language people already understand. As African communications platforms and regional PR research have increasingly shown, local media builds credibility because it is grounded rather than imposed.

And trust is what advertising is really trying to buy.

This is why the recent rush toward AI-generated media in advertising deserves more honesty than it usually gets.

Yes, AI is cheaper.
Yes, it is faster.
Yes, it can create a polished image in minutes.

But cheap creative is not the same as effective creative.

Bad ads have always been cheap, not just in budget, but in thought. They flatten complexity. They choose convenience over care and confuse efficiency with persuasion.

Advertising is not just about producing an image. It is about producing recognition.

A mother looking at an advert is not only processing lighting and composition. She is scanning for truth: Does this feel like life as I know it? Does this interaction feel earned? Does this place feel inhabited? Does this brand understand anything about the world it claims to serve?

AI can approximate aesthetics, but approximation is not the same as lived texture.

Human memory is unusually sensitive to what feels slightly off. The smile that lands too neatly. The kitchen looks perfect, but has never been used. The child’s face is expressive but somehow hollow. Even when viewers cannot immediately identify what feels wrong, they often register distance. And distance is costly in marketing.

Recent research supports this: when consumers know, or suspect, content is AI-generated, they often rate it as less authentic, less trustworthy, and less emotionally resonant, especially in categories tied to care, identity, or lived experience.

A 2025 study in the International Journal of Research in Marketing found that AI disclosure can reduce perceived trust and weaken ad attitudes in emotionally sensitive categories.

This does not mean AI has no place in creative work. It can be useful for drafts, resizing, ideation, and efficiency behind the scenes.

But when brands start replacing lived experience with synthetic shortcuts, they are not just saving money; they are stripping out the very thing that makes people care.

And in a world already saturated with content, attention is not won by making more images. It is won by making people feel that someone understood enough to notice what real life actually looks like.

For more, click HERE to join our WhatsApp channel!