More than half of Gen Zers want to be full-time influencers, a recent Morning Consult poll found, and it’s not hard to see why. They’ve grown up watching their peers on YouTube film makeup tutorials, narrate their “Fortnite” games, and explain money. It’s both where they live their lives and where they learn about the world. Plus, the career path sounds like a dream: You get to be your own boss, you can make more than you could at a traditional job, and you just have to be good on camera. Success stories of influencers making millions of dollars have only galvanized the generation.
“Gen Z explicitly wants to become influencers, are trained to become influencers, do camps to become influencers, strategize in a variety of ways to become an influencer,” Angèle Christin, a professor at Stanford studying the influencer industry, told me.
A decade ago, many successful influencers stumbled into fame; they were just regular people sharing about their lives or showcasing a niche skill when they suddenly found themselves with a huge audience and brands desperate to work with them. Anyone could strike it lucky at any time. But over the past several years, the influencer economy has shifted. Gone are the days when casual posting could suddenly turn into a lucrative career. On top of the work being an influencer requires, the competition gets tougher each day as more people vie for fame.
To make matters worse, influencer wannabes aren’t competing only with humans — they’re soon going to be competing with artificial intelligence. Advancements in AI tech have given birth to an industry of AI influencers, and major companies are beginning to show interest in their far more cost-effective approach to marketing. And who can really compete with that?
The influencer market is thoroughly saturated, Nikita Baklanov, an analyst at the influencer-marketing company HypeAuditor, said. Out of Instagram’s some 2 billion monthly active users, “only 800,000 accounts have over 100,000 followers,” he said, citing his company’s research. That’s less than 1% of accounts — but still a lot of accounts. And the number is growing.
In a crowded field, standing out takes work. Julia Broome, a social-media manager for influencers and celebrities, is already seeing the impact on longtime influencers. “Creators that were able to get a big buzz or get a huge following back in the day, they’re experiencing some drop-off,” she said.
Today, most influencers need to have a variety of revenue streams to stay financially viable, Baklanov said. They need to offer subscriptions, create their own merch, or sell a course. Increasingly, they need to exist beyond social media. “Platforms keep changing. The algorithms keep on changing. The formats keep on changing. And at some point, the influencers realize that they have to build a loyal audience, and they have to take that audience outside of social-media platforms,” Christin, the Stanford professor, told me. That means setting up a podcast, running events, or creating a newsletter.
With tools like the AI image generator Midjourney and OpenAI’s forthcoming Sora, creating content is becoming much-more affordable.
A Wall Street Journal article from 2021 suggested that the chances of making enough money from TikTok influencing for it to be an at least five-year career were less than 0.0001%. Only 100 people out of more than 1 million will succeed.
As the career path slips further out of reach, some people are turning to creating user-generated content for brands, rather than for their own audiences. Broome told me she’d seen a “big surge in UGC creators,” who come at a much-lower cost to brands than traditional influencers. For creators, it’s a nice way to make some money, but if your dream is to call your own shots for your own fan community, it’s not a good sign. And now AI is poised to be another threat to job opportunities.
In 2016, Miquela Sousa, aka Lil Miquela, was born, as a fully formed Brazilian American 19-year-old. With 2.7 million followers on Instagram Miquela has been reported to rake in over $10 million a year in brand deals, with campaigns for Prada and Calvin Klein. She, however, is not a real person.
She is one of roughly 200 virtual influencers, according to Virtual Humans, a site that tracks these faux influencers, fueled by advanced motion graphics and a sizable team of people. Along with others, like Imma, who gets millions of views on TikTok, and Shudu, a virtual model, she proved people were willing to engage with someone who wasn’t exactly real.
Miquela’s success didn’t spark a virtual-influencer revolution, but that was largely because of cost — human influencers were still cheaper. In the past couple of years, however, AI has advanced enough to make digital influencers much easier to create and run. With tools like the AI image generator Midjourney and OpenAI’s forthcoming Sora, which makes extremely realistic videos from text prompts, creating content is becoming much-more affordable. “It costs almost nothing,” Baklanov of HypeAuditor said. “And they can actually duplicate it in different languages.”
AI-generated model Aitana has 200,000 Instagram followers and reportedly makes $11,000 a month with brand deals.
The Clueless Agency
Companies have been quick to capitalize on the potential. Late last year, Meta announced “a universe of characters,” fueled by AI, with Instagram and Facebook accounts you could message with. “We’ve been creating AIs that have more personality, opinions, and interests, and are a bit more fun to interact with,” the announcement said. An AI-model agency called The Clueless launched last year with two models. The founder told Euronews, “We did it so that we could make a better living and not be dependent on other people who have egos.” Also last year, the AI-influencer company 1337 emerged from startup stealth mode with $4 million in backing. It has built dozens of AI entities each with niche interests, intricate backstories, and their own Spotify playlists.
Aurora is a 24-year-old climate-crisis activist on a mission to preserve Antarctica. Ezra, a 20-year-old intellectual studying at Oxford, likes hanging out at flea markets and cozy cafés. Wai is a 21-year-old abstract artist and sports enthusiast who’s opened an art school for young talent.
Jenny Dearing, the company’s cofounder and CEO, has been busy selecting content creators who can shape the behaviors of these entities via prompts. Once paired with a person to run things, each AI influencer will create posts and videos and start interacting with its audience. The AI learns from each interaction and uses the data that’s gathered to become more engaging. It’s no problem for an AI persona to respond to every comment or question it receives, and turn that into instant feedback for whoever is running the show.
This model of influencing requires far less time from real people, and it’s particularly suited to those who are interested in influencing but don’t want to put their own lives up for consumption. “For us, the long game is enabling more and more creators and brands to come in, create their own entity, manage that entity, advocate for that entity, and evolve and grow them over time,” Dearing said.
Does it matter that an influencer is a human using AI tools or an AI character guided by a human? Either way, they are trying to sell us something.
For brands, being able to manage a consistent and engaging social-media presence at a low cost is an attractive pitch — and some have already expressed interest, Dearing told me. She sees AI influencers being used to provide a deep level of information, support, and guidance on brands and products. “Product placement is such a tiny potential of the future,” she said. “For us, it’s really a lot of knowledge exchange.”
It will take time to see whether AI personas catch on, but the initial shift spells trouble for the future of influencing. Dearing doesn’t think traditional influencers will disappear, but she thinks a shift toward AI ones is possible. “I see both scenarios existing for some period of time,” she said. “What that balance looks like, who knows?” She added: “Over time, maybe it’s an 80-20 where brands get really excited about having that control” and human influencers make up a smaller portion of the budget.
Do AI influencers really stand a chance at building trust with an audience? Most marketing experts say that what audiences really want on social media is authenticity. How you define that is up for debate, but it’s clear that an AI-generated influencer is going to raise some eyebrows. After all, if a computer animation is promoting a new skincare product, you probably wouldn’t trust it as much as you would a human influencer swearing it was life-changing. On the flip side, though, anyone who has found themself apologizing to ChatGPT knows exactly how quickly we start to see AI as human.
A 2020 paper in the Journal of Advertising found that AI influencers “can produce positive brand benefits similar to those produced by human celebrity endorsers.” But it also found that when things went wrong with an AI influencer, there was similar reputational brand damage. There’s potential for financial harm, too: Earlier this year, an Air Canada AI chatbot gave incorrect information about a discount to a customer. A tribunal ruled that the company was bound by that and had to provide a refund. A study in the European Journal of Marketing found that consumers were just as likely to follow an AI influencer as a human influencer but that they didn’t trust the AI influencer as much. They were, however, more likely to talk with others about the AI influencer, which could turn out to be a positive for brands.
The trust question may already be moot, though. As AI chatbots have become more commonplace, most of us are already using them regularly. Chatbots are used for healthcare support, for therapy, and to warn teens of the dangers of too much social media — how different is an AI influencer, really? Was the Air Canada customer wrong to trust the AI bot’s information? Clearly, the court didn’t think so.
You could even argue that there’s something more authentic about a brand using an AI entity to market itself and engage with people than a human turning themself into a brand to appeal to the algorithmic robots of a social-media platform. We already know that influencers carefully craft their presentation, often with a manager or coach. Does it matter that an influencer is a human using AI tools or an AI character guided by a human? Either way, they are trying to sell us something.
The social-media-management platform Hootsuite’s 2024 Social Media Trends report said: “The most successful brands will redefine ‘authenticity.’” The focus won’t be on who is creating the content but on whether the content is compelling. As Dearing sees it, we’re moving toward an information-driven experience. “Ultimately, these influencers become a really lovely visual way to engage on social platforms that are more knowledge-based,” she said.
Some experts have a significantly more pessimistic take: Eric Schmidt, a former Google CEO, and Jonathan Haidt, a social psychologist and author, wrote last year about the potential for “skillful manipulation of people by AI super-influencers.” This would be possible, they suggested, by the potential for generative AI to be highly personalized to individual wants, needs, and interests. AI has also been used to create deepfakes of celebrities and influencers that can damage their reputations. Already, there’s plenty of interest from marketers in AI’s ability to exploit consumer cognitive biases. Essentially providing marketers with a highly targeted, souped-up version of the most effective human influencers.
If AI does catch on in the influencer world, Gen Z and Gen Alpha will have an even more challenging time striking influencer gold. What holds value online is already starting to undergo a fundamental shift, and influence will be up for grabs.
Clem De Pressigny is a freelance writer and editor, and was previously the editorial director of i-D magazine. She covers the internet and technology, the climate crisis, and culture.