AI Influencers Vs. Real Humans: What Actually Works In Marketing?

Authored by Karla Rossouw 

First, it was “Is It Cake?” where viewers were staring at the television, trying to figure out whether they were looking at a cake or an everyday object. Now, social media has its own version of the guessing game. Except this time, it isn’t about what’s on the plate, it’s about who’s on the screen.

AI influencers are no longer just strange-looking digital characters that appear in the occasional marketing campaign. Some have millions of loyal followers and audiences that willingly follow their fictional lives. Others have been designed to look so realistic that it can be difficult to tell where the human ends and the AI begins.

You might think that real human influencers would have an advantage over AI ones. After all, why would anyone take advice from someone who doesn’t even exist? But what if that assumption is actually much further from the truth? Recent research is starting to fill the gaps when it comes to what makes an influencer successful and some of the findings are probably not what you’d expect.

Can People Really Tell If an Influencer Is AI?

A 2024 study published in Technology in Society put this to the test. The researchers created Instagram accounts representing AI influencers and ordinary people of the public. They then asked 43 participants to evaluate these accounts. Participants had difficulty distinguishing between the AI and human-owned accounts, while AI-generated content was rated similarly in quality to content created by human influencers.

That doesn’t mean people are completely fooled by AI, but it does debunk the old assumption that viewers would immediately recognise artificial content. The bigger issue may therefore be what happens after someone realises an influencer isn’t human.

Can Knowing It’s An AI Influencer Change the Reaction?

In July 2026, 2,012 UK adults were asked about AI-generated actors in advertising. The results weren’t particularly enthusiastic. 69% of people said that they’d seen online adverts that featured AI-generated actors, while 76% of those who had seen them said they didn’t like them. Another 68% referred to AI-generated actors as less convincing than human actors.

Research also found that 78% of Britons would be less trusting of a brand that used AI-generated actors and this number increased to 83% among 18- to 34-year-olds. The fact that consumers are comfortable seeing something artificial on their screens doesn’t necessarily mean they’re happy being marketed by it.

Something similar was revealed in the UK Harris Poll, in its 2026 research into AI-generated advertising. Seven in ten consumers said they trusted adverts created entirely by humans. Six in ten said a person should always be involved when AI is used in advertising, while 79% said brands should disclose their use of AI.

The Research Doesn’t Show Humans Winning Everything

If the story ended there, the answer would seem pretty obvious: stick with real people. But the research paints a picture that is a bit more complicated.

A 2026 study published in the Journal of the Academy of Marketing Science looked at hundreds of experiments comparing AI and human influencers. Interestingly, it found that AI influencers aren’t necessarily worse at getting people to engage with content. In fact, they performed just as well as human influencers for consumer engagement. However, they were less effective when it came to credibility, and getting people to act on their recommendations.

And that difference is important for brands. Getting someone to stop scrolling because an influencer looks unusual is one thing, but winning trust and persuading people to buy products is a completely different ballgame. So, where an AI character might win the first battle, a human creator who genuinely uses and talks about a product could still have that competitive edge.

Some Of the Biggest AI Influencers Don’t Even Look Human

If looking like a real person were the secret to succeeding as an influencer, you’d expect the most realistic characters to dominate the industry, right? That’s not what the research shows.

Look at Nobody Sausage, the infamous 3D-animated influencer known for its strange noodle-legs, colourful pink body, and physics-defying dance moves on social media. This fictional character reportedly has more than 7.3 million Instagram followers and a 4.34% engagement rate. Additionally, John Pork, another fictional character, is reported to have more than 500,000 followers and a 9.42% engagement rate.

So, how are they performing so well? The research links their good performance to the type of content they produce, including humour, relatable lifestyle content and memes, rather than their ability to look like actual human beings. Nobody follows a sausage because they genuinely believe it is a person. The appeal is the character itself. In other words, an AI influencer doesn’t necessarily need to fool anyone to work. So, the question shouldn’t be whether AI influencers can replace humans. Instead, it should be whether the creator actually fits the type of content being produced.

What Real Influencers Still Bring to The Table

The truth is, human influencers still have something that AI creators can’t bring into their content. Humans have real experiences.
For instance, an influencer who struggles with Psoriasis could talk about the physical discomfort, feelings of isolation and the negative impact it has on their confidence. They make their audience feel seen, they build connections and gain trust. If that same influencer does a product review and speaks about how it gave them some relief, they could persuade others to try it out as well.

Don’t get me wrong, this doesn’t mean that every human influencer is trustworthy, but it certainly gives their advice a foundation that an AI persona doesn’t naturally have.

Humans or AI? Which One Should Brands Choose?

For a brand selling something where personal experience is central to the purchase decision, a real creator may still make more sense. Beauty, travel, fitness and lifestyle products often depend on someone being able to demonstrate genuine experience. Would you really listen to the weight loss journey of an AI influencer and buy protein powder from them?

For campaigns built around entertainment, fantasy, gaming or futuristic products, this may actually place human creators at a disadvantage because of human limitations. For instance, a virtual character can exist in a world that doesn’t follow normal rules or become part of a story that would be difficult to build around a conventional influencer.

Cost and control must also be kept in mind. Research on AI influencers tends to focus on the advantages around creative freedom, consistency and fewer geographical constraints. A brand doesn’t have to organise flights, work around a creator’s schedule or worry about whether the influencer would suddenly change direction.

The Influencer Doesn’t Have to Be Human, But the Strategy Does

If there is something brands could take from the research, it is that they shouldn’t abandon human influencers or rush to create AI ones in abundance. Appearance alone isn’t what determines whether an influencer can lead a successful marketing campaign.

Factors like entertainment, relevance, character and context can all drive engagement. For instance, credibility becomes more important when the audience needs to believe the person behind the recommendation. And when AI is being used in advertising, UK consumers increasingly expect brands to be open about it.

This means that instead of choosing between a real person and an artificial one, brands need to ask what role the influencer is actually playing in the campaign. Sometimes the right person will be human, sometimes it might be a cartoon sausage with millions of followers and sometimes the most successful campaign could use both.