For years social networks have sold us an illusion: that of authenticity. Seemingly ordinary people who show us what clothes they wear, what cosmetics they buy, what restaurants they visit or what mobile phone they carry in their pocket. Advice on travel, nutrition and sports given by people who could be our neighbor. That closeness was precisely the secret of the influencers’ success. They seemed real people sharing real experiences. But in just over a year, something is changing.
More and more brands are replacing content creators with characters generated entirely through artificial intelligence. They are attractive, photogenic young people, always available and capable of producing content twenty-four hours a day without getting tired, without making mistakes and without asking for salary increases. The problem is that many times the public doesn’t even know that they don’t exist.
According to an analysis of Guardiansome companies already use AI-generated influencers to promote products on Instagram, TikTok and other platforms without clearly informing that they are characters created with artificial intelligence. In fact, in some cases, Confidentiality agreements are even used between the agencies and the client to hide the use of these tools.
The question is inevitable: Does it really matter if the person recommending a product is human or not? The idea is not new. One of the best-known cases is that of Lil Miquela, a virtual influencer created in 2016 who has accumulated millions of followers and has collaborated with brands such as Calvin Klein or Prada. What started as a technological curiosity has become a multi-billion dollar industry.
For agencies and brands, an influencer made with AI does not get sick, does not generate controversy and cannot work for the competition. It can adapt to any campaign, speak any language, change its appearance in seconds and maintain an image perfectly aligned with the brand’s values. To this we must add that it is becoming easier and faster to create these characters. What once required entire animation studios can now be done from a laptop.
However, the main strength of influencers was always the feeling of authenticity. A study published in the Journal of Marketing & Social Research analyzed the reaction of 320 users to promotional content created by human influencers and artificial influencers. The results showed that the AI-generated characters had significantly lower scores (more than 60%) in perceived authenticity and trust towards the brand. Even more striking: when participants were explicitly informed that the influencer was artificial, trust decreased even further. The paradox is evident.
Companies want transparency, but transparency can reduce the effectiveness of the message. A Cornell University study warns that generative tools are dramatically lowering the barriers to creating fake content at scale. It is not just about producing realistic images, but about building complete identities capable of publishing, interacting and generating credibility for months or years. In other words, we are not dealing with simple automated ads. We are faced with characters capable of developing a parasocial relationship with thousands or millions of people.
The European Youth Portal points out that many virtual influencers share similar characteristics: they are extraordinarily attractive, lead seemingly perfect lives and project impossible beauty standards. The difference is that they are not even limited by human biology. Its creators can modify any physical feature with a click.
“On the legal level – notes the EU statement – in mid-May of this year, the Member States developed the first regulations for the use of AI, focusing on mandatory labeling and the protection and safeguarding of fundamental rights. However, there is already a clear gap between legislation and practical experience. This is largely due to the rapid development of artificial intelligence.” In short, not only is it late, but those responsible for legislating assume it.
Meanwhile, the platforms They try to introduce labels that identify images or videos generated by AI. The problem is that these marks can disappear when the content is re-edited or distributed outside of official channels.
But the most important problem is not technological or legal, but psychological. AI profiles quickly and negatively affect the mental health of adolescents. And to this we must add that Before we wondered if a photograph was manipulated or if a video was fictitious, now we question if a person is real. Or if the product you are promoting even exists. And when we can no longer confidently answer those questions, social media’s most valuable asset, trust, could also become its most scarce.