Meta Platforms generates a significant amount of revenue from advertising.
Meta reported third-quarter total revenue of $51.2 billion, a 26% year-over-year increase. Of that total, about $50 billion came from advertising.
And AI is only making advertising on Facebook, Instagram, and WhatsApp more precise.
“Our ads business continues to perform very well, largely due to improvements in our AI ranking systems,” CEO Mark Zuckerberg said during the company’s third-quarter earnings call. According to Zuckerberg, AI ranking systems have driven both “better performance and efficiency.”
But the company isn’t using AI just to optimize its back end; it’s also using large language models to improve engagement and its recommendation systems.
With all that attention and computing power directed toward advertising, it would make sense for Meta to also improve its issues with scam advertising.
However, a new report reveals that Meta not only acknowledges its scam advertising problem, but also earns billions from it.
Meta CEO Mark Zuckerberg claims AI is enhancing the user experience on Meta Platforms, but internal documents suggest a different perspective.Bloomberg/Getty Images
If you’ve ever come across a Facebook or Instagram ad for an AI-generated product, such as a “Nostalgia Bus Slow Cooker,” for an incredibly low price, you are not alone.
According to a series of leaked internal documents viewed by Reuters (subscription required), Meta exposes its 3.5 billion platform users globally to approximately 15 billion scam advertisements a day.
These “higher risk” ads show clear signs of being fraudulent, but Meta also earns about $7 billion in annualized revenue from them every year, according to the documents.
Related: Veteran analyst says Meta stock is stuck in ‘near-term purgatory’
In fact, Meta internally projected in 2024 that it would earn about 10% of its overall annual revenue (about $16 billion) from running ads for scams and banned goods.
Meta’s internal warning systems flagged much of the marketing fraud; however, the company only bans advertisers if its automated systems predict that the marketers are at least 95% likely to be committing fraud, according to the documents.
If the advertisers don’t reach that threshold, but the company still thinks it’s likely that they are scamming, Meta will just charge higher ad rates as a penalty.
Not only is AI not helping the problem, but it’s actively making it worse — the documents note that users who click on scam ads are likely to see more of them, due to Meta’s aforementioned AI-powered ad-personalization system.
The documents Reuters viewed span a timeline from 2021 to 2025 and include files from across Meta’s finance, lobbying, engineering, and safety divisions.
According to the news service, these documents reflect Meta’s efforts to “quantify the scale of abuse on its platforms,” but they also reveal the company’s limited efforts to address the issue.
Related: US credit card holders spent much less on this luxury in October
Meta did not immediately respond to a request for comment, but told Reuters that the documents in its report “present a selective view that distorts Meta’s approach to fraud and scams.”
The company also referred to its own 10% scam revenue estimate as “rough and overly-inclusive.” According to the company, it has reduced user reports of scam ads globally by 58% and has removed more than 134 million pieces of scam ad content this year.
Remember that the company hosts about 15 billion scam ads per day, according to its own estimates.
Dominick Miserandino, a retail expert with three decades in ad tech, offered TheStreet some perspective.
A presentation in May 2025 from the company’s safety staff estimated that Meta platforms were involved in a third of all successful scams in the U.S., and it acknowledged in other internal docs that some of its main competitors were doing a better job at eliminating fraud on their platforms.
While some of its rivals are fighting the scam wave, some industry experts argue that the big tech companies are simply not doing enough.
“I’ve spent 20+ years on every side of the digital ad ecosystem: agency, publisher, and tech. This book has been read before,” Adverge.ai founder Andrew Moskowitz told TheStreet.
“None of this surprises me,” he added. “A portion of digital media spend has always been questionable. The industry needs to refocus on trust and transparency, supported by real governance that ensures accountability.”
With 3.5 billion users across its platforms, Meta reaches more people than any other network in history.
It is a goldmine for scam artists, and advertising scams are just one of the many types of frauds being perpetrated on the platform.
“Users should approach all social media interactions [on Facebook], whether it is a post, a tweet, or a direct message, with a healthy dose of skepticism,” Eva Velazquez, CEO and president of Identity Theft Resource Center, recently told Reader’s Digest.
Ragebait: Scammers intentionally produce rage-inducing content to drive interactions.
Giveaways: Nothing in life is free, not even that iPad you could win by signing up on a Facebook page you don’t follow.
Facebook Quiz: Many of those quizzes are just gathering info on you to sell to advertisers — some fish for security-question answers, like your mother’s maiden name.
Fake Coupon Codes: If the code requires you to give up personal information, just pass.
Facebook Marketplace: Don’t pay or communicate outside of Facebook; Facebook Purchase Protection only covers payments made through Facebook Checkout. Source: Reader’s Digest
Related: $75 billion company to pay millions in remote work dispute
This story was originally reported by TheStreet on Nov 7, 2025, where it first appeared in the Technology section. Add TheStreet as a Preferred Source by clicking here.
