Covers of bestselling books in the sloppy style of AI scammers Layer Ø This story comes from The Hollywood Reporter‘s upcoming AI Issue, out in April.
Authors are at their most vulnerable in the weeks leading up to the publication of a book. With little to do but wait for reviews and social media reactions, they vacillate between the manic hope that the fruit of their labor will imminently top the bestseller list and the dread that it will fall between the cracks of the attention economy and disappear into the void, which, after all, is the fate of most books. One of my fellow authors ruefully calls this nerve-wracking pre-pub period “the calm before the calm.”
Related Stories
The AI Issue As Hollywood Panics Over AI Job Losses, a Startup Says It Has the Answer
Business The Sora-Disney Collapse: What Does It Mean?
It was in this harrowing lull ahead of the January release of my latest book, a nonfictional nautical tale titled Neptune’s Fortune, that I began receiving the emails. Sent by purported publishing professionals with names ranging from insipidly Anglo Saxon (Dorothy Stratton) to extra-terrestrial (Futa Concept), they all followed a similar template: Four or five smoothly written paragraphs, beginning with detailed, excessively flattering descriptions of my book — e.g. “Neptune’s Fortune is a masterclass in historical adventure and human obsession” (oh shucks) — and culminating in an offer to help me increase its visibility. They promised to mount social media campaigns, get me on podcasts, flood Goodreads with positive reviews, or introduce me to online book clubs, the new gateways to viral success these days as traditional media wanes in influence. Not mentioned in that first pitch was the fee they would charge for these services. But follow-up emails would cite estimates ranging from hundreds to thousands of dollars — a small price for glory.
To get this out of the way, this is not one of those stories where the writer admits to having been swindled out of their life savings. I admit I took the flattery at face value the first time I read it, but by the time money was mentioned, it was obvious it was a scam. So obvious that I couldn’t imagine anyone who’s ever been online actually falling for this, let alone enough people to justify the effort it takes to tailor a detailed pitch letter to each individual mark. Given the average writer’s income, I couldn’t imagine the juice to be worth the squeeze. How could this scam possibly be worth it for the scammers? Enter AI. These emails instantly gave off the sickly-sweet reek of LLM sycophancy, all written with the same bland fluency. Some included links to hastily assembled Potemkin websites featuring AI-generated covers of non-existent books by non-existent authors who had supposedly availed themselves of the sender’s company’s services. Artificial intelligence seemed to infuse every aspect of these pitches, down to the sender’s profile photo, typically of a comely young woman at a nondescript office or coffee shop, with occasional nightmarish hallucinations in the background. Whoever was behind this was not paying for the premium image generators.
Before long, I was receiving several of these a week. The polished, personalized, idiomatic language seemed genuine enough to evade gmail’s spam blockers. I thought I had the rare misfortune of landing on a sucker list, but I soon learned that almost all of my peers were suddenly experiencing the same deluge. Some were receiving thousands. The AI productivity boom we’ve been promised may not yet have hit the economy at large, but it’s turbocharged this kind of small-bore scam, drowning our inboxes in highly customized slop.
The emails began appearing in inboxes in mid 2025, and fame and success have provided no immunity. Patrick Radden Keefe, the bestselling author of Say Nothing and Empire of Pain, both of which were adapted into hit TV shows, says it’s non-stop. “Every morning I wake up to two or three of these emails,” he tells me. Dan Brown, whose The Da Vinci Code has sold more than 80 million copies, shared one such missive on Facebook: “Hey Dan Brown, I just came across [The Da Vinci code, and I’ve got to say it’s rare to see a story that balances heart, message, and craft the way yours does. It’s the kind of book readers don’t just read: they feel it. That’s special.” Now here came the pitch: “But here’s the thing: too many incredible books like yours disappear in the noise.”
The technology’s versatility could theoretically allow scammers to target any creative profession. They’ve begun going after musicians with a similar flattery-forward approach, but evidently to a far lesser degree than writers.
Authors, with our particular stew of ego and insecurity, have long been prime targets for scammers. A few years ago, before the widespread use of chatbots, an elaborate operation based out of the Philippines allegedly bilked some 800 mostly self-published authors out of a total of $44 million dollars, promising to turn their books into movies and charging exorbitant fees for services that never materialized. But those were more sophisticated stings. The advent of free generative AI has allowed for a shift in scam strategy, from targeted sniper attacks to a hail of machine gun fire. It has brought the cost of bullets — in this case, the time and effort it takes to find a mark, research their work, draft an email, and keep a conversation going — down to essentially zero. Even if the vast majority miss their target, a single hit keeps the con going. A numbers game is easy to win when the numbers are limitless.
It’s hard to say how many authors are being taken in, since victims can be embarrassed to admit they were duped. “I have been badly scammed by somebody named Moses Fred,” confessed one commenter on Writers Beware, a site devoted to exposing such swindles. “No website, no LinkedIn page, no following. I was a total fool. He kept emailing me and telling me he could do wonderful things with my book and I fell for it and I feel like a fool – a fool!”
Victoria Strauss, the founder of Writers Beware, told me she has personally spoken to fewer than twenty authors who have paid the scammers. Even if that represents just a fraction of the people who’ve fallen for the ploy, I would have expected more, given the onslaught of daily emails sent to tens of thousands of professional authors in America, to say nothing of the self-published ones and those abroad. But thanks to the speed and facility with which AI allows emails to be drafted, it might well be enough. It’s likely paying off or else the emails would have slowed rather than increased.
The evangelists of generative AI told us it would unleash imaginations. “I believe AI has the potential to democratize creativity on an unprecedented scale,” argued Mira Murati, OpenAI’s former Chief Technical Officer. The technology, we were told, would yield a limitless artistic bounty by removing the barriers to entry for anyone lacking in opportunity (or talent). We’re still waiting for that future, but in the meantime, it’s certainly removed barriers for scammers.
***
“There are no new scams, honestly. Only new versions of old scams,” says Edward Balleisen, professor of history at Duke and the author of Fraud: An American History from Barnum to Madoff. The AI literary scam, he says, is merely a contemporary variant of an old con often called the “Spanish Prisoner” or, more generally, the “advanced fee scam,” in which the fraudsters dangle a large reward in exchange for a small upfront payment. Once the payment is made, the scammers either disappear or ask for just a little more. “The basic nature of the scam is there’s a bait and switch,” Balleisen says.
It’s the nature of LLMs not to invent anything new but merely remix — and cheapen — what exists. What they’ve done is amplify the scammers’ reach. Says Balleisen, “AI is permitting people who would not have had this sort of familiarity with the [publishing] market or with the lingo to be able to pass themselves off as somebody who could offer those services.”
As susceptible to flattery as professional authors can be, they’re also hard to fool, since they earn a living from manipulating words. “They’re always talking about my books in this sort of pseudo-knowledgeable way,” says Keefe of the scammers, “and they all just regurgitate the same language. And it’s sort of fascinating, right? Because, if you’re in the business of actually stringing together original thoughts, you realize what they’re doing is averaging things. I think most [of us] are pretty quick to see through that.”
For authors, the insult of being targeted by an AI scam adds to the injury of knowing that the AI attempting to steal from them has already stolen from them, since it was trained without permission on their collected works. As the technology muscles in on everyone’s turf, it’s led to a flurry of lawsuits and surge of public opposition. Authors were among the first on the barricades, alongside Hollywood screenwriters who made resistance to AI central to their 2023 strike. (Disclosure: I was the lead plaintiff in a class action against OpenAI and Microsoft, the first such lawsuit by non-fiction authors to go after generative AI for copyright infringement. The suit is ongoing at the Southern District of New York, and though I’ve since yielded my place among the plaintiffs to more prominent colleagues, my objection to having copyrighted work used without consent or compensation remains steadfast.)
After getting enough of these emails, I decided to answer a few to get a better sense of how they worked and who was behind them. I picked one at random: Nilda Mulan, whose AI-generated profile photo bore a passing resemblance to, yes, Mulan, if she wore pantsuits. She had told me that Neptune’s Fortune was “a truly captivating work,” and that she wanted to feature it in The European Book Club, which would introduce it to “our growing global community.” Naturally, this wouldn’t be free: “To support the quality, coordination, and promotion of each feature, participation includes a modest participation fee, which covers editorial preparation, audience outreach, and dedicated promotional efforts.”
I told her that I was interested, but that, given the number of scams out there, I would need to verify that she was who she said she was. If she could show me a picture of herself holding my book, which she claimed to love, or even a newspaper with today’s date, I’d consider doing business with her. I felt conflicted about toying with her, knowing that the people engaging in this kind of scam are often themselves victims of human trafficking and have little choice in the matter. But I felt that the only way to peek behind the AI was to start a conversation.
Nilda’s five-paragraph response, which hit my inbox less two minutes later, deftly deflected my request as only AI could: “Thank you for your candid message. I completely understand your caution, and I appreciate your desire to verify legitimacy before moving forward,” she wrote, before informing me that sharing photos was “against our internal policies.” We went back and forth like this for a while. Nilda’s responses remained unctuously flattering and polite, but never gave in. It felt like speaking with a chatbot instructed to rebuff me, because that’s effectively what it was.
In lieu of the photo I’d asked for, Nilda sent me a link to The European Book Club’s website, to assure me that it was a legitimate entity. Among its list of recently featured books were several literary classics with misspelled or otherwise mangled covers that bore all the hallmarks of cheap, over-the-counter AI. What, you’ve never read Murakami’s Norrvgigian Wood or García Márquez’s Ona Hun dr Da de Solitude?
Having hit a wall, I stopped responding. Five minutes later, eager to keep me on the hook, Nilda apparently made an exception to her internal policies and sent me the below image of her avidly reading my book, coincidentally wearing the exact same outfit and sitting in the same chair as in her profile photo.
Courtesy When I asked what was up with the garbled text on the book’s spine and back cover, Nilda, or more likely her LLM of choice, told me it was “simply due to the image quality and compression, which can distort small details like spine and back cover text in photos.”
I didn’t know what I was hoping to achieve, exactly. To embarrass the scammer into admitting their wrongdoing? As long as they were hiding behind their chatbots, this game could go on forever. To make any progress in my investigation, I’d have to speak to someone face to face, or as close as I could get.
I picked out another sender, “Nicole Powell,” whose profile photo showed her to be a young, perfectly lit white woman smiling impishly behind her laptop. She had offered to help me promote my book on podcasts and in newsletters, among other outlets. I told her that my publisher, Crown, was helping me with all those things, but that I would be interested in speaking with her over Zoom to discuss other opportunities.
Nicole agreed, to my surprise, but said her assistant, Penny, would attend the meeting in her stead. When I insisted on speaking with Nicole — meaning the woman in Nicole Powell’s photo — if only for two minutes, she responded, “I completely understand, and I appreciate your preference. Due to my current schedule, I will not be available for calls until further notice, so my assistant is the one who can attend at this time. She is fully briefed and able to address any questions you have.”
And so I joined a Zoom with Penny. The woman in front of me, who was almost certainly the person passing herself off as Nicole, looked to be in her early twenties. She was shabbily dressed and appeared to be missing one of her bottom teeth. The wall behind her was filthy and I could hear voices yammering loudly in the background. Penny wouldn’t make eye contact as she spoke, with far less fluidity than her AI, about helping me establish an “organic promotion plan” for social media.
It was a sad spectacle. I reminded myself that many cyber-scammers are effectively enslaved. In developing countries around the world, trafficking rings lure desperate job seekers — including children — overseas with fake listings, confiscate their travel documents and force them into fraud operations targeting people in richer countries. The recruits are given unattainable sales quotas to meet, often on penalty of physical torture, quotas that AI has presumably driven higher by allowing scammers to cast a wider net. I had no way to know if this was Penny’s case. I wouldn’t have gotten an honest answer if I’d asked her. But based on the apparent helplessness of the person in front of me, it seemed more than plausible. (Penny is not the name she gave me, which may or may not have been her real name; I’m giving her a pseudonym anyway to protect her from potential punishment.)
She told me Nicole was busy for the foreseeable future, but that I could speak with her instead. “You can tell me the things you want.” When I told her that sounded a bit scammy, she said, “If we are a scam, I won’t be able to get on a video call with you.” OK then.
Though Nicole had assured me Penny was fully briefed on my book (which she’d called “a gripping blend of historical investigation, maritime adventure, and imperial history”) Penny seemed to know nothing about it, or me.
There was something raw about seeing the human being behind the scam without her AI interpreter. However duplicitous, she seemed completely at a loss, naked. I couldn’t imagine this poor girl convincing anyone that she might know the secrets to literary success in America. I wondered who was more credulous: the people who fell for this stuff, or Penny, for thinking she could fool anyone without the help of AI. Just as overreliance on ChatGPT for homework is leading to a loss of literacy among high school students, swindlers are growing so dependent on LLMs they’re incapable of swindling without them.
While she spoke, I Googled her full first name and surname, as displayed on her Zoom tile. Both are of Yoruba origin (they respectively translate as “person of wealth” and “wealth has arrived”) and are common in Nigeria. I asked where her company was based. After a short hesitation, she said South Africa. Perhaps that was true. Or perhaps she knew that the notorious Nigerian Prince scam had given the country a bad name, even if it didn’t actually originate in Nigeria.
This one, apparently, did. Strauss, the founder of Writers Beware, says most of the scammers request payment on platforms like Paypal, Upwork, and Fiverr, as well as Coachly, which is popular among West African freelancers. Following a complex trail of IP addresses and geolocated Twitter posts by the scammers, Strauss claims to have traced the operation primarily to Nigeria.
“They’ve become unbelievably numerous and prolific,” Strauss says of the emails. Yet she believes the Nigerian ring itself to be relatively small. “AI makes this volume possible,” she says. “The generative AI platforms are making it possible to write endless exchanges.”
***
The scams are targeting the book business at an uncertain time for the industry, when self-publishing, rebranded as “indie publishing,” is on the rise, and traditional marketing methods — newspaper reviews, book tours, radio appearances — are proving ever less effective. A book’s success is increasingly dependent on the unpredictable force of online virality, an intimidating prospect for many authors who don’t have vast social media followings, aren’t sure how it all works, and would rather not have to deal with it.
“BookTok, especially for a lot of new authors who aren’t super invested in social media, it’s kind of a mysterious place,” says Christie Hinrichs of Authors Unbound, an advocacy and events agency for authors. “Like, how do you even find an entry point in TikTok? How do you go viral? These questions could definitely be a point of insecurity that [the scammers] are taking advantage of.”
Hinrichs has another theory for why the scam is so prevalent and presumably effective to some degree: authors feel like publishers aren’t holding up their end of the bargain. “I don’t know how people in Africa would know this on an industry level, but post-Covid we have seen publisher support around new titles withdraw a little bit, and some authors might feel underserved, in which case I could see scammers capitalizing on that fear or misgiving.”
Most publishers have pages on their websites warning against scams impersonating their staff or otherwise targeting authors. But they struggle to keep up with the pace, as new variations on the scam emerge every day. “It’s a whack-a-mole situation,” says the general counsel for a major publishing house who requested anonymity. “I email my contact at the FBI to ask what we’re supposed to do about this. But they just refer me to the IC3.gov website,” referring to the Internet Crime Complaint Center. Ultimately, the AI lit scam is just one drop in the ocean of scams perpetrated by criminal rings across the planet, a problem too intractable and widespread for any one country’s authorities to solve.
All effective scams are built on desire. So, too, is the book industry. For Elyse Graham, a bestselling historian who shared her frustrations about the AI scam with me, the desire that authors of all levels share — for success, recognition, prestige — is up there with the other major sources of yearning that scammers take advantage of: “It’s love, it’s get-rich-quick, and it’s book publishing: those are the three things where the engine of longing is so great that it can power an entire scam infrastructure,” Graham says, semi-jokingly. (Hollywood dreams, it could be argued, are just as powerful an engine, as evidenced by the self-published authors who fell prey to the Filipino scammers dangling Netflix deals.)
The hunger of authors “is something like the hunger that gets a 90-year-old to believe that they’re in a relationship with Keanu Reeves,” Graham says, in reference to a widespread AI-enabled celebrity-impersonation scam originating from Southeast Asia.
The lit scammers are also impersonating celebrities, albeit of the bookish kind. Some of their emails purport to be fan letters from superstar authors like Margaret Atwood and Rick Riordan. These titans don’t make a pitch directly. The goal is to hook the recipient into a conversation, in the course of which “Riordan” or “Atwood” or “Stephen King” will presumably let the desperate author in on the secret organic promotion plans that have enabled them to conquer the bestseller list.
Literary celebrities are invoked in other ways, too. When I asked one scammer (“Becky Janice”) to connect me with authors she had worked with, she provided me with contact information for three writers, including none other than Elena Ferrante, the pseudonymous Italian author behind the so-called Neapolitan novels, the global phenomenon that was adapted into the HBO series My Brilliant Friend. (I can reveal here, exclusively, that her email is apparently [email protected].) Ferrante’s name, it turns out, is frequently cited in these emails, perhaps because the scammers know that Ferrante’s true identity remains elusive.
Less prominent authors have also been impersonated. After I’d been given a false email address for Adam Makos, the bestselling military historian, I contacted the real Makos to let him know his identity was being stolen. He was already well aware of the scams and how often they were using his name.
“When I realized it was widespread,” he says, “my concern was, what if someone actually believes this? This could be crushing to a new author’s morale. It’s easy to laugh off the Nigerian scams at this point. But it’s different when you start playing with someone’s emotions — not just their bank account — their emotions, their outlook on humanity. Their belief in the future and their belief in their own potential.”
Makos had alerted his local sheriff’s office, to no avail. He is now planning to contact the FBI but fears any attempt to rein in scammers on the other side of the world is bound to fail. A better solution, he thinks, would be to regulate the US-based AI companies these operations depend on, but that, too, seems futile given the Trump administration’s laissez-faire attitude towards AI. “Usually you would look to the government, but I feel like our legislators are kicking the can down the road. They’re going to let some other generation worry about this,” he says. “I think what we’re watching right now in real time, is the death of truth, the death of reality. Where are the safeguards?”
There is a life cycle to scams, says Balleisen, the fraud historian. Once enough people are targeted, press reports start coming out, awareness spreads, and people grow wise. Nobody but the least informed, or most gullible, among us would now fall for the Nigerian Prince scam, for example. When the scammers’ hit rate has gone down low enough that it’s no longer worth their while, they move on to another scam. (Of course, by allowing such a vast volume of outreach, AI has greatly lowered the bar for that minimum hit rate.)
When Balleisen told me this, I felt virtuous about having done my part, by writing this article, to expose the scam, and thankful that the benighted legions of striving authors had someone as AI-savvy as me to protect them from this new threat. My pride didn’t last long. While I was putting the finishing touches on the piece, I received an evite from a friend in L.A. to celebrate his birthday the following weekend. I clicked the link and found an image of him blowing out candles. I entered my phone number and sent my regrets, telling him I wouldn’t be in town. Turns out the image was AI and his birthday wasn’t for months. Ever since then, my email has been flooded with AI evites from “friends.” There’s a sucker born every minute.
THR Newsletters
Sign up for THR news straight to your inbox every day
Subscribe Sign Up
