A popular pro-Trump Instagram persona named "Emily Hart", who posted pro-life, pro-Second Amendment content and racked up millions of views, never existed. She was fabricated entirely by artificial intelligence, operated by a 22-year-old medical student in India who told Wired he was "making good money" off conservative Americans he openly mocked as gullible.
The New York Post reported on the Wired investigation that unmasked "Sam," an orthopedic surgery trainee who built the Hart persona from scratch using AI image generators, then monetized it through merchandise sales and paid adult content on Fanvue, an OnlyFans competitor.
The episode is a case study in how cheaply and quickly a foreign operator can build a fake political identity, attract a real American audience, and extract real American dollars, all while laughing about it.
Sam told Wired he got the idea while looking for ways to earn money during medical school. He wanted to save enough to emigrate to the United States after graduation. He turned to Google's Gemini AI for advice, and the software reportedly told him "the conservative audience (especially older men in the US) often has higher disposable income and is more loyal."
So he built Emily Hart. Her profile said she was a registered nurse. Her AI-generated face had what Sam described as "Jennifer Lawrence looks." Her content hit every note designed to attract conservative engagement: pro-Christian, pro-gun, anti-abortion, anti-immigration. One post showed her firing a rifle. Another read: "If you want a reason to unfollow: Christ is king, abortion is murder, and all illegals must be deported."
The account gained 10,000 followers within a month. Every reel Sam posted garnered millions of views, Wired reported. He was soon raking in thousands of dollars a month.
Sam described the daily effort as minimal. He told Wired he spent 30 to 50 minutes a day on the operation, posting content he described in blunt terms:
"Every day I'd write something pro-Christian, pro-Second Amendment, pro-life, anti-abortion, anti-woke, and anti-immigration."
The merchandise arm came next, MAGA-themed T-shirts. Then Sam created a Fanvue account, where paid subscribers could access AI-generated images and messages. He used Grok AI to generate nude photos of the fictional Hart and uploaded them to the platform. Fanvue, unlike some competitors, allows AI-generated content.
The convergence of AI and political manipulation is no longer theoretical. It is a functioning business model.
What makes this story sting is not just the deception. It is the operator's open contempt for the people he was deceiving. Sam did not pretend to feel conflicted. He told Wired exactly what he thought of his audience:
"The MAGA crowd is made up of d*** people, like, super-d*** people. And they fall for it."
He contrasted them with Democrats, whom he said "know that it's AI slop, so they don't engage as much." Whether that assessment is accurate or self-serving, it reveals the cynicism at the core of the operation. Sam was not an ideologue. He was a foreign grifter who identified a market and exploited it.
He bragged about the ease of it all:
"I was basically doing nothing. And it was just flooded with money."
And yet, when pressed, he offered a familiar dodge. "I don't feel like I was scamming people," Sam told Wired. That claim is hard to square with the facts: a fake identity, fake photos, fake political convictions, and real money flowing from real people who believed they were supporting a real woman.
Instagram took down Hart's profile in February for "fraudulent" activity. The platform requires creators to disclose whether their content is AI-generated, a policy Hart's account apparently violated. A Facebook account linked to the persona remained online even after Wired published its investigation, and was only removed afterward.
The timeline matters. Hart's account had time to build a massive audience, generate millions of views, and pull in steady revenue before any platform intervened. By the time Instagram acted, the damage, financial and reputational, was already done.
Valerie Wirtschafter, a fellow at the Brookings Institution who studies emerging technology and democracy, told Wired that AI has supercharged the old problem of fake online profiles:
"AI has made them more believable, and there has perhaps been an amplification of it."
Wirtschafter also noted that the conservative female persona was especially effective because "most women ages 18 to 29 skew liberal," making a young, attractive, outspoken conservative woman "more attention-grabbing", and, by extension, more profitable.
The growing use of AI-generated content for exploitation, whether political, financial, or sexual, is drawing increasing scrutiny. A recent federal conviction under the Take It Down Act showed how synthetic media can be weaponized in far darker ways. The Hart case sits on the grift end of the spectrum, but the underlying technology is the same.
The left will use this story to mock Trump supporters. That is predictable and irrelevant. The real lesson is about information hygiene, and it applies to everyone, but especially to a political movement that prides itself on rejecting mainstream media narratives.
If you are going to distrust legacy outlets, you need sharper filters for the alternatives, not weaker ones. A persona with no verifiable biography, no public appearances, and no traceable history should raise flags, regardless of how perfectly her posts align with your beliefs. Alignment is exactly what a con artist manufactures.
Sam understood this. He built Hart to be a mirror, reflecting back every value and grievance his target audience held. The content was not persuasion. It was flattery dressed up as politics.
Congressional oversight of tech platforms continues to lag behind the pace of the threat. Lawmakers like Senator Chuck Grassley have launched probes into tech giants for failing to police harmful content, but the enforcement gap remains wide. Instagram's disclosure requirement for AI content is meaningless if it only catches fraudulent accounts after they have already harvested an audience and revenue.
Sam told Wired he would have stopped posting Hart content on his own. Maybe so. But the infrastructure he used, Gemini AI for market research, Grok AI for image generation, Fanvue for monetization, remains available to anyone with a laptop and a willingness to lie.
The broader collision between artificial intelligence and public trust is playing out across multiple fronts, from controversies surrounding AI industry leaders to courtroom battles over how the technology is deployed. The Hart case is a small-dollar grift in the grand scheme, but it exposes a vulnerability that scales.
The financial loss to individual followers was probably modest, a T-shirt here, a Fanvue subscription there. The deeper cost is trust. Every fake account that gets exposed makes it a little harder for real conservative voices, real nurses, real veterans, real women who hold these beliefs, to build audiences without suspicion.
Sam pocketed the money and plans to leave India for America. The people he deceived are left to wonder who else in their feeds is a fiction. And the platforms that hosted the fraud moved only after journalists did the work for them.
The MAGA movement has real people, real energy, and real grievances. It does not need AI-generated mascots built by foreign operators who hold it in contempt. The best defense against the next Emily Hart is not better AI detection software. It is the old-fashioned habit of verifying who you are actually listening to.
A con only works when the mark stops asking questions. That is a lesson no algorithm can teach.
