Meta internally estimated that approximately 10% of its anticipated revenue for 2024—around $16 billion—would stem from advertising linked to scams and prohibited products, according to internal company documents.
Reviewing previously undisclosed documents obtained by Reuters, it appears that the social media giant failed for at least three years to identify and prevent a surge of ads exposing billions of Facebook, Instagram, and WhatsApp users to fraudulent schemes involving e-commerce, illegal online casinos, and the sale of banned medical supplies.
In December 2024, one internal report estimates that the platform displays around 15 billion high-risk scam ads daily—ads exhibiting clear signs of fraud. This category of scam advertising generates approximately $7 billion in annualized revenue for Meta, the document states.
Most of this fraudulent activity originated from marketers flagged by Meta’s internal warning systems, yet the company’s policy only bans advertisers if its automated filters are at least 95% confident they are engaged in fraud. For ads flagged by the system but not reaching this certainty, Meta imposes higher ad rates as a penalty, aiming to discourage such suspect advertisers from continuing to place ads.
Furthermore, the documents highlight that users clicking on scam ads are likely to encounter additional similar ads due to Meta’s ad personalization algorithms, which tailor content based on individual user interests.
The confidential documents, created between 2021 and 2024 across various divisions—including finance, lobbying, engineering, and safety—reveal Meta’s continuous struggle to measure the scale of abuse and its hesitance to implement stricter measures that could harm its business interests.
Sandeep Abraham, a fraud specialist and former Meta safety investigator turned consultant, commented that Meta’s acceptance of revenue from potentially fraudulent sources exposes a lack of rigorous oversight in the advertising industry. “If regulators wouldn’t tolerate banks profiting from fraud, then they shouldn’t allow it in tech,” he stated.
Meta spokesperson Andy Stone responded by asserting that the documents present a skewed and narrow view. He described the internal estimate of 10.1% of 2024 revenue from scams as “rough and overly inclusive,” explaining that subsequent assessments showed the figure to be lower because the original estimate included many legitimate ads. He declined to specify a revised percentage but emphasized that the company’s efforts to combat fraud are substantial.
“Meta actively fights scams because users, legitimate advertisers, and the platform itself all dislike this content,” Stone added. “Over the past 18 months, we’ve reduced scam ad reports globally by 58% and removed over 134 million scam ads so far in 2025.”
Some internal documents indicate Meta plans to intensify its efforts in 2025, aiming to cut scam ads in certain regions by up to 50%. Managers also congratulated teams on successful reductions of scam content.
However, internal research suggests Meta’s platforms have become central to global fraud networks. A May 2025 presentation estimates that a third of successful scams in the U.S. occur on Meta’s platforms, and the company acknowledged that competitors like Google are more effective at rooting out fraud.
A 2025 internal review concluded it’s easier to run scams on Meta than on Google, though it didn’t specify why.
Around this time, regulators worldwide are stepping up pressure. In the U.S., the SEC is investigating Meta for running ads related to financial scams, and in the UK, regulators found that in 2023, Meta was involved in 54% of all payments-related scam losses—more than twice the combined total of other social platforms.
Meta’s CEO Mark Zuckerberg testified before the U.S. Senate Judiciary Committee in January 2024 amidst these mounting regulatory concerns. The company’s internal documents indicate that addressing the issue of scamming often involves weighing the costs of enforcement against potential fines. For instance, the documents reveal plans to reduce the contribution of scam ads to overall revenue from an estimated 10.1% to 7.3% in 2025, with further reductions targeted in later years.
Despite efforts to curb illegal activity, the company recognizes that penalties for scam ads—up to $1 billion—are trivial compared to the billions earned from such ads each year. As of late 2024, Meta earns approximately $3.5 billion every six months from scam ads presenting higher legal risks, such as those misrepresenting brands or individuals.
Rather than proactively policing ads, internal discussions suggest that Meta tends to respond mainly when faced with imminent regulatory action. As part of its strategy, Meta has also experimented with increasing costs for suspected scam advertisers through “penalty bids,” where high-risk advertisers must pay more at auction to place ads, thereby decreasing their profitability.
Following the rollout of this approach, initial tests showed declines in scam reports and a slight drop in overall ad revenue. The company also continues to limit the amount of revenue it is willing to risk by acting against suspicious advertisers, setting internal thresholds equivalent to roughly $135 million in the first half of 2025—about 0.15% of total revenue—though a company representative clarified this isn’t a strict cap.
In October 2024, amid pressure, senior executives proposed a more cautious enforcement approach, focusing on countries where regulatory action was imminent rather than a broad crackdown. This led to targets aiming to reduce scam-related revenue from 10.1% in 2024 to roughly 7.3% by the end of 2025, and further down to 5.8% by 2027.
Reports show that in 2022, Meta uncovered a network of accounts falsely claiming to be U.S. military members in conflict zones, which sent millions of fraudulent messages weekly. The company also noted a rise in sextortion scams targeting teenagers, along with a flood of fake celebrity profiles and brand impersonations worldwide.
Despite the increase in online fraud, Meta’s internal assessment from 2022 described scam ads as a “low severity” issue, mainly considering them a poor user experience. Focus was primarily on impersonation scams involving well-known figures and brands, which posed a threat to advertiser relations and revenue.
Layoffs and resource reallocations between 2022 and 2023 hindered enforcement efforts. Staff focused on brand concerns were let go, with many safety team resources diverted toward virtual reality and AI projects. As a result, the company largely neglected to address the growing tide of scam activity, with some internal documents revealing that up to 96% of user-reported scams were ignored or mishandled.
Meta’s safety team aimed to improve this response, intending to process a higher percentage of valid scam reports moving forward. For example, in 2023, the team aimed to lower the rejection rate of legitimate scam reports from 96% to no more than 75%.
Erin West, a former prosecutor now leading a nonprofit against online scams, criticized Meta’s general approach of ignoring user reports, noting that genuine reports often go unaddressed.
An incident involving a Canadian military recruiter’s Facebook account highlighted these issues. After the account was hacked and impersonated, Meta was slow to respond, and the scammer used the profile to promote fake crypto investments. Despite multiple reports from the user and her contacts, Meta took about a month to take the account offline. This delay resulted in other military colleagues falling victim to scams, with some victims losing thousands of dollars.
Authorities struggled to recover the stolen funds, which had been transferred to bank accounts in Nigeria, out of reach for enforcement due to jurisdictional restrictions. Meta declined to comment specifically on this case.
Meta classifies scams into “organic,” meaning unpaid fraudulent activities like fake marketplace listings or charlatan health groups, as well as “paid” scam ads. The company estimates it faces about 22 billion organic scam attempts daily, on top of the 15 billion paid scam ads shown to users each day.
The platform’s current policies often miss much of this activity. For example, Singaporean police provided Meta with 146 scam examples last fall, only to find that less than a quarter violated the site’s policies, with some violations falling short of the letter of the rules but still representing deceptive practices.
Other flagged scams, like crypto ads claiming to feature the Canadian prime minister, were not caught under existing policies, although Meta later removed them for other reasons. Internal documents acknowledge significant gaps in the platform’s scam detection policies, highlighting the challenge in closing these loopholes.
Even when advertisers are caught, enforcement can be lenient. A 2024 report states that small advertisers engaged in financial fraud typically have to be flagged multiple times—up to eight—before they’re shut down. Larger accounts, known as “High Value Accounts,” may evade restrictions after more than 500 violations.
Some anti-scam campaigns have reached enormous sizes, with four campaigns alone responsible for $67 million a month in ad revenue before being removed. To further combat scams, Meta has experimented with increasing the costs for suspected rogue advertisers through “penalty bids,” which require high-risk advertisers to bid more to place ads. This strategy aims to make scams less profitable and reduce the number of scam ads shown.
Following the implementation of this system, Meta reported declines in both scam reports and overall ad revenue, with the company aiming to strike a balance between financial interests and platform safety.
Source: REUTERS
Created with Gemini
