FYI: Your Favorite Influencer Might Be AI

Fake people are social media's next frontier.

Baddie Betty is an 82-year-old influencer with 745,000 followers on Instagram. Her bright silver hair is her signature, and it’s typically styled with a side part and flipped ends. Betty’s known on the platform for her no-nonsense dating advice (she’s even written a book about it) and flaunting her wealthy lifestyle. In one photo, she’s lounging on a bed—showing off her red bottoms, of course—surrounded by designer shopping bags. In another picture, Betty is sitting on a pilates reformer, alongside her pink Hermes bag, wearing a matching Alo set and holding an iced matcha. On her story highlights, under “PR,” there are photos of products she has allegedly been gifted by brands like Mermade Hair, Anomalia Paris, Ballerette, and Saltair.

The thing about Betty, however, is that she is not real. She’s AI. (According to Meta’s policies, you’re required to disclose if you’re posting AI-generated content, but it’s not clear if the company is actually enforcing these rules.) I noticed that a friend had liked one of Betty’s posts, so I texted her: “...you know she’s AI right??” She had no idea.

Lil Miquela first appeared on Instagram in 2016, and she’s widely considered one of the first virtual influencers. (In her bio, she discloses that she’s a “robot.”) She’s worked with brands from Prada and Dior to BMW and Samsung. She’s even released music. It’s unlikely that someone would come across one of Lil Miquela’s photos and mistake her for a real human being. She looks like a hyper-realistic Sims character.

This newer generation of digital influencers, however, are harder to detect, thanks to improving technology. It’s not difficult to imagine someone scrolling through their Explore page and coming across content from an AI influencer without ever discovering that they’re not real.

This isn’t just worrying for content creators who are worried about their livelihoods, but for the rest of us, too, who are navigating this warped reality, trying not to get scammed by a deep-fake of a reputable dermatologist, and just want to be able to trust their own eyes.

Why Are AI Influencers Becoming More Popular?

AI influencers have no personal likes or dislikes, no opinions, and no thoughts of their own. That’s part of their appeal, at least for some brands and companies. There’s no worry that an influencer will go rogue or say something off-script and alienate part of their audience. They can be programmed and controlled by their creator. In fact, one might say that there’s little difference between an AI influencer and a billboard. According to Business of Fashion, most brands are still hesitant at the idea of partnering with AI influencers, though that hasn’t stopped agencies from creating them. The Clueless, an agency based in Barcelona, has a whole division dedicated to AI models. Arguably, their most famous creation is Aitana López, “Spain’s first 100% AI-generated influencer,” according to their website. Lopez shares videos of her applying lipgloss and posts gym selfies even though she, technically, can’t exercise. She’s shared a sponsored post of her skincare routine, collaborated with Amazon on a Black Friday post, and has a partnership with Spanish hair salon, Llongueras.

While some AI influencers, like López, are managed by agencies, there are several created by individuals. A trio of influencers—Bailey Moore, Fiona Pellegrini, and Emily Pellegrini—are the work of a man by the name of Professor EP (his Instagram has since been deleted). Professor EP claims that he’s made $1 million in less than six months from these accounts. However, even with all of this alleged income, he sells an “AI Influencer Accelerator” course where he promises to show customers exactly how to “build and monetize” their own AI influencer. The faces of these AI influencers come from real people. Wired reported that in one of his guides, Professor EP recommends creating the faces of these AI influencers using real people: “If, for example, you like Ariana Grande’s eyes and Kylie Jenner’s lips, you can create an image which combines these attributes by including these specifications in your prompt.” Professor EP also teaches people how to make deepfake face swaps using real people’s content: “Face-swapping content from other accounts without permission seems to work well for a lot of AI influencers,” read the instructions on one of his videos. Then, according to Wired, he recommends setting up a Fanvue account for your AI influencer and talking to “lonely people” to sell explicit, AI-generated content.

It’s unclear how well Professor EP’s guide is selling, though one buyer seems to be unhappy with his purchase. Still, Reddit is filled with posts, like this one, where users detail how they make thousands of dollars with AI influencers, thanks to affiliate links and Fanvue accounts. Most of the people behind these female AI influencers appear to be men.

With AI influencers in general, though, it’s difficult to pinpoint the exact person behind the account. This means that men could be perpetuating harmful beauty standards or a white person could use black avatars to create rage bait content. This could happen to any one of us—our likeness can be used online by someone else without any meaningful repercussions.

AI influencers don’t just have no likes, dislikes, or opinions; they also don’t have any human needs. They don’t need to file invoices, sleep, eat, rest, or shit. They can spend all of their time churning out content for the algorithmic machine. Take Sacha Mjarvis, an AI fashion influencer on TikTok with almost 100,000 followers, as an example. During a recent 24-hour period, I counted that she had posted 22 videos. Most human influencers post around one to three videos on TikTok a day.

Meta and TikTok require a label on AI-generated content, but many influencers including Mjarvis, don’t have one in their bio. Laws around the use of AI differ depending on what state you live in. In December, Governor Kathy Hochul of New York signed the nation’s first law requiring companies to disclose when they use AI-generated avatars in ads. However, there still isn’t any federal legislation regulating AI in the United States.

How AI Impacts Beauty Standards

Around 15 years ago, patients would bring photos of celebrities, like Angelina Jolie, to Dr. Goretti Ho Taghva’s office and say, “I want her nose,” or “I want her eyes.” Then patients started bringing in photos of influencers. Now, people bring in AI-generated images of themselves. The problem is that people don’t realize just how unrealistic these optimized images are—even more so than FaceTuned selfies—or how far they are from what real human faces actually look like. “They don’t realize that what they’re looking at is a digital fantasy,” explains the plastic surgeon. “AI distorts bone structure, texture, so much. It can’t be surgically replicated.”

In the past celebrities and social media have shaped beauty standards, and now the proliferation of AI-generated images is impacting them, too. AI reproduces already existing beauty standards, and it doesn’t challenge or pushback on norms. “The margin for what’s considered ‘acceptable’ has shrunk dramatically as the algorithm’s beauty standards have converged into one homogenized face,” says Dr. Goretti, “and that’s the standard of beauty that’s being fed into these machine learning systems.”

When you’re used to seeing these heavily filtered and AI-generated images constantly on your feed, that becomes your reality. A real human face with pores, zits, and wrinkles becomes abnormal. “It’s extremely harmful,” Dr. Goretti adds. “People just can’t deal with looking at themselves in the mirror. No matter how beautiful they are, they only see those ‘faults,’ the pores, pigments, and wrinkles. They don’t realize that that is just what a normal face looks like.”

Who Is All Of This For?

AI companies promised that this technology would usher in a new “Intelligence Age” that would lead to the discovery of a cure for cancer, an end to climate change (ironic given generative AI negative impact on the environment), and less monotonous work for employees. Instead we’ve gotten an overvalued search engine, hallucinating chatbots, AI slop all over social media, increased carbon emissions, higher energy bills, and greater job insecurity, including for content creators.

It’s not clear that people even want AI influencers; once people discover that an account is AI, they’re called out in the comments or in separate posts. A 2025 Pew Research survey found that half of US adults are more concerned than excited by regular AI use, and 76% said that it’s important to be able to tell the difference between AI- and human-generated content.

The reason influencers rose in prominence in the first place was because they were real. It was a departure from the sanitized, corporate content of the past. They shared their honest opinions and gave insight into their everyday lives, the highs and the lows, which created a connection and built trust between them and their audience. Why would someone knowingly buy a serum or blush recommended by someone who doesn’t even have real skin? AI may be able to sell a perfect image but they can’t build community or trust, two things that set humans apart from the robots.

—Ama Kwarteng