What an exploratory sample of fifteen Flemish students teaches us about AI, trust and brand communication.
Criticism of AI is growing louder. The Flemish newspaper
recently described how resistance is emerging particularly among Gen Z towards a technology that tech companies often present as inevitable. That resistance isn’t only about privacy, copyright or the impact on jobs. The feeling that AI is being imposed on people, and that human creativity increasingly has to make way for efficiency, also plays a role.
At the same time, young people are enthusiastic users of AI tools themselves. They use ChatGPT for schoolwork, search for information with AI and encounter generated texts, images and videos every day. Use and enthusiasm are therefore not necessarily the same thing.
That tension is relevant for brands too. Young people don”t appear to be categorically opposed to AI content. Their main questions concern how companies use the technology. Is AI being used to support people or to eliminate as much human involvement as possible? Has the content been checked, and is it reliable? Or does it feel generic, misleading and lazy?
Today, brands use AI for brainstorming, texts, images, translations and videos. That is hardly news anymore. The more interesting question is how your target audience views the result.
Does it matter to young people whether a text was written with the help of AI? Do they trust a brand less when it uses AI-generated images? And should you always disclose that AI was involved in the production process?
Mattia Gielens explored these questions during his internship at Mediaforta. For his bachelor’s dissertation at UCLL, he conducted in-depth interviews with fifteen Flemish students aged between 18 and 25.
Fifteen interviews are, of course, not a representative reflection of all young people in Flanders. We therefore can’t draw definitive conclusions from this sample. Nevertheless, several patterns emerged remarkably often. They offer valuable food for thought for brands and content marketers.
All participants used AI tools themselves. They also assumed that companies use this technology.
The use of AI in itself therefore prompted little outrage. Young people don’t seem to expect every text to be written entirely from a blank screen or every image to be created through a traditional photoshoot. But that doesn’t mean they accept every use of AI.
The most important boundary concerns human involvement. Above all, the respondents wanted to retain the sense that someone takes responsibility for what a brand publishes. A person must check the content, verify the facts and decide whether the result is good enough.
Almost all the participants believed they could recognise AI content. They felt particularly confident about images and videos.
Among other things, they looked for unnatural hands, perfect faces, bright colours, strange body proportions and movements that were not quite right. Some images were described as “too good to be true”.
Opinions were more divided when it came to text. Some participants believed they could quickly recognise ChatGPT’s typical structure. They pointed to formal wording, predictable lists, an overly polished structure and the absence of a personal tone.
Other participants found it difficult to distinguish AI-generated texts from texts written by people.
One of the clearest findings was that tolerance of AI varies considerably depending on the type of content.
For blogs, newsletters and other written content, the young people generally had fewer objections to AI support. The condition was that the information was correct and that a person had checked the final text.
There was greater sensitivity around images. AI-generated people in particular prompted resistance. When a brand replaces real employees, customers or models with non-existent people, it is more likely to create a sense of deception.
The subject also plays a role. For financial, medical or news-related content, participants expected far more human oversight than for a light-hearted social media post.
Twelve of the fifteen young people interviewed considered it important to know whether content had been created by a person or by AI. Even so, they didn’t all advocate a general disclaimer for every text or image.
Here too, context proved decisive. Many participants considered transparency important for realistic AI-generated images, factual information or sensitive subjects. They saw less need for a prominent AI label on a simple advertising text or newsletter, particularly when AI had only been used as a supporting tool.
The respondents reacted most negatively when AI was conspicuous, careless or constantly present.
Excessive use was associated with laziness, cost-cutting and a lack of effort. That is an important reputational risk. Your target audience may feel that you are asking for their attention while investing as little as possible in the communication yourself.
A single AI-assisted text is unlikely to make or break a brand. A continuous stream of generic blogs, unnatural images and predictable social media posts may do so.
Moreover, quality isn’t assessed solely on a publication-by-publication basis. People recognise patterns. When every text follows the same structure and every image has the same polished appearance, an artificial brand world emerges in which nothing feels real anymore.
The respondents did not set the bar equally high everywhere. On TikTok and Instagram, there was more room for experimentation, entertainment and striking AI effects. On a company website, they expected greater professionalism, reliability and oversight.
As a result, the same content may feel creative on one channel and damage your credibility on another.
That doesn’t mean anything goes on social media. Misleading images or artificial people can prompt resistance there too. It does mean that the purpose and environment of the channel help determine what young people consider acceptable.
