Earlier this year, Emily Schuman, the founder of the beloved lifestyle blog Cupcakes and Cashmere, was scrolling through Instagram when she stopped cold. There she was—her face, her smile, her familiar tousled hair—standing in a sponsored ad, holding a tiny vial of GLP-1 medication and promoting Gala, a telehealth company she had never heard of. The emily on the screen looked exactly like her, but she wasn’t her. It was a digitally fabricated version of an old selfie she had taken in her car years ago, manipulated with artificial intelligence to make it seem as though she was endorsing a product she had never tried, never discussed, and never agreed to promote. Before she could even process what she was seeing, her followers had already noticed. Messages began flooding her inbox: Did you really promote that? Is that you? Are you okay? The ad felt completely out of character with her brand, and her audience knew it. But the worst part was that it wasn’t a one-off mistake. The same stolen likeness began popping up everywhere—an AI version of her with darker hair applying foundation for a skincare brand she had never contacted, and TikTok clips featuring generated images of her endorsing a blood-testing company called Superpower. Emily later described the experience as deeply violating. The images looked just like her, she said, but like “the upside down version” of herself. She had spent years building her reputation, curating a genuine connection with her followers, and doing honest work with brands she believed in. Now, in a few clicks, strangers had hijacked her identity and turned it into a marketing tool.
Emily’s experience is not some rare, freakish incident. In the age of generative AI, it has become a terrifyingly common workplace hazard for anyone whose face is their livelihood. Influencers are among the easiest people to impersonate because they live their lives in public, generating thousands of photos and hours of video that an AI model can study to produce a convincingly lifelike fake. Researchers who study this new wave of digital fraud say that deepfakes have become so good that they are nearly indistinguishable from reality, and the financial damage adds up fast. One investigation estimated that consumers around the world may have lost as much as $3.7 billion to deepfake scams in just one year, with social media impersonations accounting for about half of that total. Celebrities like Taylor Swift, Oprah Winfrey, Kim Kardashian, Gwyneth Paltrow, and Tom Hanks have all been swept up in these schemes, their faces pasted onto fraudulent ads for crypto schemes, questionable supplements, and knockoff luxury goods. But ordinary working creators are just as vulnerable, if not more so, because they don’t have the armies of lawyers and publicists that A-listers can call to make problems disappear overnight. Beauty influencer Arielle Lorre discovered that her face had been used to promote a skincare brand she had never tried. Fashion creator Jessi Caparella found altered versions of her photos on Pinterest, redirecting unsuspecting shoppers to affiliate clothing links that had nothing to do with her. OnlyFans creator Elaina St James watched as deepfaked videos of her were used to convince fans to send money to fraudulent accounts. In each case, the emotional toll was heavy. It is one thing to lose a brand deal or see your engagement dip; it is another to lose control of your own face, to have it twisted into a version of you that you never were and never wanted to be.
The rise of these fake endorsements is fueled, in part, by a broader shift in the advertising world toward AI-generated content. Marketers have always been in the business of illusion, airbrushing imperfections, touching up photographs, and creating perfect worlds with clever editing. But generative AI has changed the rules entirely. Now, brands can produce polished marketing content from nothing, without paying for a photoshoot, a creative team, or even a real human model. Virtual influencers—digital avatars designed to look like people—have become surprisingly successful. According to one analysis, AI-generated or virtual influencers accounted for around $1.37 billion, or roughly 4% of all brand spending on influencer marketing, in a single year. They are cheap, tireless, infinitely malleable, and always on message. They never age, never get tired, and never cause scandals—at least, not in the way that real people do. For brands, that makes them incredibly attractive. But this enthusiasm for AI has blurred the line between legitimate digital creativity and outright theft. Instead of inventing an entirely new virtual face, some companies have quietly taken the faces of real influencers and put them into manufactured advertisements, scraping years of public content to create what looks like a genuine endorsement. It is a massive betrayal of the trust that influencers have built with their audiences, and it threatens to destroy the entire foundation of the creator economy. After all, the whole reason influencer marketing works is that followers believe a real person is sharing a product they genuinely love. When that authenticity is counterfeit, the entire system begins to crumble.
Nowhere was the gut-wrenching violation of this practice more vivid than in the ordeal of Molly Tranchin, a fashion influencer known online as FashionVeggie. Molly, who has more than half a million followers, signed a contract with an underwear brand called Eby to create three marketing videos in exchange for a fee. According to her lawsuit, she completed her end of the bargain, filming the videos in a way that fit her personal brand. In the footage she submitted, she posed modestly, carefully covering her chest and breasts with her arms, exactly as she and the brand had agreed. But instead of posting her work, the brand allegedly used AI to create a deepfake video of Molly in a completely different pose, exposing her breasts and nipples through one of its sheer bras. It was not a subtle edit or a minor tweak; it was a wholesale reimagining of her body and her consent. The video made it appear as though Molly had agreed to appear in a sexualized ad she had never seen, with a level of exposure she had never authorized. The legal system has so far offered little in the way of relief. Molly was forced to drop her lawsuit because of a question about which court had jurisdiction, and she plans to refile in a different venue. But the damage has already been done. Her face, her body, and her professional identity were commandeered in the most intimate possible way, and even if she eventually wins in court, the memory of that fake video will linger online like a ghost she can never quite outrun.
At first glance, it might seem as though the platforms where these deepfakes appear should be the ones to solve the problem. After all, Meta and TikTok have both said explicitly that they prohibit the use of artificial intelligence or digital manipulation to replicate a person’s identity. But policies on paper have not translated into protections in practice. Emily Schuman and her manager reported the fake GLP-1 ad to Meta immediately, and so did many of her followers. It still took weeks of relentless badgering before the company finally removed it. The fake makeup ad, meanwhile, was still circulating on Instagram despite repeated requests to take it down. The truth is that these platforms make money from engagement, and outrageous content, including fraudulent ads, drives clicks. They also allow advertisers to use hyper-targeted audience slicing, which means bad actors can quietly launch a scam, reach a highly specific audience, and then start all over with a new account whenever one gets flagged. The platform’s own systems are often used against the very people they claim to protect. Experts who study this issue are not optimistic. They point out that Meta makes an enormous amount of money from ads promoting scams and banned goods—one investigation estimated that such ads accounted for roughly 10% of the company’s revenue, about $16 billion in a single year. An internal presentation from Meta’s own safety team reportedly estimated that the company’s platforms were involved in a third of all successful scams in the United States. It is hard to believe that a company which profits so handsomely from this ecosystem feels an urgent need to dismantle it. The platforms have introduced new ad verification measures and an AI detection tool aimed at stopping celebrity impersonation, but these efforts are often reactive, slow, and demonstrably insufficient.
For Emily Schuman, the fight has become a painful part of everyday life. She keeps discovering new fakes, new stolen images, new fabricated endorsements, and she is now weighing whether to hire legal counsel simply to help her chase down the endless stream of content and file takedown notices. It is an exhausting, expensive, and ultimately deeply unfair position to be in. She didn’t ask for this. She didn’t do anything wrong. She just built a career in a world that increasingly rewards visibility, and now that visibility is being weaponized against her. Every hour spent fighting a deepfake is an hour lost from creating, engaging with her audience, and doing the actual work she loves. Every dollar spent on lawyers is a dollar that could have been invested in her business. But the alternative is even worse—letting the fakes multiply until they permanently tarnish her reputation. As she put it, it’s a lose-lose situation from every angle. And this is the uncomfortable truth that her story reveals. The threat of deepfakes is not just a problem for celebrities and influencers. It’s a warning to all of us who share our lives online, indeed to anyone who has ever posted a selfie. If the face of someone with half a million followers can be stolen and manipulated so easily, what chance do the rest of us have? The creator economy promises that with enough hard work and authenticity, anyone can build a meaningful career out of who they really are. But in this strange new world, the technology that can build your brand can just as easily steal it. Until platforms take real responsibility, until laws catch up with the speed of AI, and until brands reject the shortcut of digital theft, people like Emily are left to chase ghosts across the internet, fighting to reclaim a face that should have been theirs all along. The deepfakes keep coming, and behind every one of them is a real person who trusted the digital world enough to show their true self to it—only to have that truth twisted into a lie.