Noah Romsdal Hallundbæk Sørensen
Leaked internal documents reveal how Meta has been profiting from online scams for years – and just how much the company has made from the deception.
Meta internally estimated late last year that about $16 billion of its 2024 revenue — roughly 10% of the company’s total — would come from ads promoting scams or banned goods, according to documents reviewed by Reuters.
Those files describe years of struggle by the company to control a flood of deceptive ads across Facebook, Instagram and WhatsApp. The documents say Meta’s platforms deliver around 15 billion high-risk scam ads to users every day.
Despite internal warnings, the company continued to profit from these campaigns, earning an estimated $7 billion a year from that category alone.
High-risk ads
According to Reuters, the documents show that much of the fraudulent activity came from advertisers flagged by Meta’s systems as suspicious. But the company only bans an advertiser if it’s at least 95% sure the ad is fraudulent. If suspicion is lower, Meta instead charges a higher price to discourage them from buying more space.
That means users who click on scam ads often see more of them, thanks to Meta’s personalization algorithms.
Screenshots reviewed by Reuters show false ads impersonating public figures and well-known brands, including spice maker McCormick and the White House.
Meta’s internal reports indicate that for years, it failed to stop the expansion of fraudulent advertising despite being aware of the problem.
Internal resistance
The confidential documents, spanning from 2021 to 2025, reveal a tension between Meta’s financial goals and its safety responsibilities.
Executives calculated how much illicit ad revenue the company could afford to lose while minimizing regulatory penalties.
Former Meta fraud investigator Sandeep Abraham told Reuters, that the company’s behavior showed a lack of oversight in the online advertising world. “If regulators wouldn’t tolerate banks profiting from fraud, they shouldn’t tolerate it in tech,” he said.
Meta spokesman Andy Stone rejected the notion that the company profited knowingly from scams, calling the internal revenue projection “rough and overly-inclusive.”
He told Reuters, that the company later revised the estimate downward and insisted that Meta “aggressively fights fraud and scams.”
Pledges to improve
Stone said Meta has reduced global user reports of scam ads by 58% in the past 18 months and removed more than 134 million pieces of scam content in 2025. A 2024 internal memo set goals to reduce scam advertising by up to half in key markets this year.
However, other internal presentations acknowledge that Meta’s products still play an outsized role in online fraud. One May 2025 briefing estimated that Meta’s platforms are involved in one-third of all successful scams in the United States.
“It is easier to advertise scams on Meta platforms than Google,” a 2025 company review concluded, though it did not specify why.
Rising scrutiny
Meta now faces inquiries from regulators on both sides of the Atlantic. According to internal records, the U.S. Securities and Exchange Commission is investigating its handling of financial scam ads. In Britain, a regulator said Meta’s platforms accounted for 54% of all payment-related scam losses in 2023 — more than twice the combined total of other social networks.
As Meta invests tens of billions in artificial intelligence and infrastructure, CEO Mark Zuckerberg has told investors the company can fund its ambitions from advertising revenue. “We have the capital from our business to do this,” he said in July.
Still, internal discussions show concern that cutting off fraudulent advertisers too abruptly could hurt Meta’s earnings. The company estimated future fines related to scam ads could reach $1 billion — far less than its scam-related revenues.
Guardrails on losses
According to Reuters, documents from early 2025 reveal that Meta’s enforcement teams were told not to take actions that could cost the company more than 0.15% of its total revenue — around $135 million in the first half of the year. “We have specific revenue guardrails,” one manager wrote.
Stone said that figure was a projection, not a fixed limit. But internal strategy papers show Meta executives approved only a gradual reduction in scam-related revenue — from 10.1% in 2024 to 7.3% by the end of 2025, with further cuts planned through 2027.
Scams and layoffs
The company’s problems with fraudulent accounts date back years. In 2022, Meta uncovered a massive network of fake profiles posing as U.S. soldiers to defraud users. Other schemes involved impersonations of celebrities and brands.
At the time, Meta labeled scam ads as a “low-severity” issue and cut back on teams working on ad safety during company-wide layoffs. Staff handling brand-rights complaints were let go, and safety units were told merely to “keep the lights on.”
By 2023, users were filing around 100,000 valid scam reports a week, but Meta ignored or wrongly dismissed 96% of them, internal data shows.
When scams hit home
Reuters spoke to a Royal Canadian Air Force recruiter whose hacked Facebook account was used to promote a fake cryptocurrency investment. Despite dozens of reports, Meta took weeks to act — and at least four of her colleagues lost money.
Canadian police traced some of the stolen funds to Nigerian bank accounts, but recovery proved impossible. “People were being harmed because they trust me,” the recruiter said.
Meta declined to comment on the case.
Policing fraud
A December 2024 presentation said Meta users face 22 billion “organic” scam attempts daily — unpaid posts, marketplace listings, and fake profiles — on top of the 15 billion paid scam ads.
Meta’s rules often fail to catch misleading promotions. When Singaporean police provided 146 examples of local scams, Meta found only 23% violated its policies.
Others — such as fake concert tickets or counterfeit job offers — were deemed within policy “in letter, but not in spirit.”
Even accounts impersonating political leaders could slip through. One 2025 document flagged $250,000 in crypto scam ads from an account claiming to be Canada’s prime minister, which Meta’s systems initially failed to detect.
The ‘Scammiest Scammers’
Meta staff began circulating weekly “Scammiest Scammer” reports highlighting advertisers with the most fraud complaints. But being featured wasn’t always enough to get them banned.
Some high-spending advertisers amassed hundreds of strikes before being suspended. Internal files show four fraudulent ad campaigns brought in $67 million a month before removal.
Meta also introduced a “penalty bid” system, forcing suspected scammers to pay higher prices in ad auctions. The aim, the company said, was to make fraudsters less competitive — though it also increased Meta’s revenue from those ads.
Stone said testing showed the system reduced scam reports, even as overall ad revenue dipped slightly.
Sources: Reuters
This article is made and published by Noah Romsdal Hallundbæk Sørensen, who may have used AI in the preparation
