According to internal documents cited by Reuters, Meta projected that around 10% of its revenue It could be linked to irregular or illegal advertising, a figure that, in business terms, would have been around 16.000 million in the exercise analyzed.
The papers describe an ecosystem in which users are exposed to about 15.000 billion “high-risk” ads per daywhile the company would only ban accounts when its systems reached a 95% certainty of fraud; in scenarios of lesser conviction, the response was to increase the advertising bid of the suspect.
What do the internal documents say?
Reuters reports that Meta's confidential self-assessment, prepared between 2021 and 2025 by finance areas, Engineering and security quantified the impact of fraudulent and prohibited product advertisements, with an internal estimate of 10,1%. of revenue in 2024 and close to 7.000 million associated with formats of “higher legal risk”.
Part of the material suggests that the company weighed the cost to suppress that content in the face of potential regulatory sanctions: executives viewed the loss of revenue from removing it as “comparable” to fines potential, and prioritized actions in regions where the sanctions would be more severe.
The review also includes targets to significantly reduce these practices: a cut of up to 50% in certain markets and a trajectory that would reduce the weight of these revenues to around 5,8%. in the medium term.
How the penalty works and why it matters
When the expulsion threshold was not reached, Meta activated the calls “penalty bids”Advertisers with signs of fraud had to pay more to win the auction, making their campaigns less profitable and, in theory, reach fewer users.
The documents describe rules stricter for small customers (blocking after multiple violations) and greater tolerance with accounts of high value, even capable of accumulating hundreds of notifications without losing access, an asymmetry that fueled internal and external complaints.
The impact on revenue, according to the cited evidence, was mixedFewer prints were being sold to suspected fraudsters, but the price increase offset some of the drop, while reducing—though not eradicating—the overall exposure to fraud.
Impact on users, authorities and market
The advertising personalization system can exacerbate the problem: those who click on a scam often see More similar adsreinforcing the risk. In recent months, campaigns have been detected with AI-powered impersonations of figures like Warren Buffett or Elon Musk and, in the Hispanic world, of affordable presenters in Spainto attract investment scams.
In the United States, the Attorney General of New York, Letitia jamesHe coordinated 42 prosecutors to demand from Meta greater controls on investment announcements on Facebook. Meanwhile, Reuters reports that the SEC has focused its attention on the spread of potential financial scams.
In Europe, a regulator of United Kingdom It attributed 54% of its 2023 payment fraud losses to Meta products, more than double that of other social media platforms, which has intensified the regulatory pressure in the region.
Furthermore, documents cited by the agency indicate that the company ignored or improperly rejected up to 96% of reports valid user reviews during certain periods, although Meta claims to have rectified these review processes.
Meta's response and the changes underway
Spokesman Andy Stone maintains that the reports offer a “selective vision” He said the 10,1% figure was “approximate and too inclusive” because it included “many legitimate ads.” He did not provide an updated figure but defended the progress made.
According to the company, in the last 18 months a 58%. User reports of fraudulent ads globally have increased, and so far in 2025, there have been removed more than 134 million of suspicious ads.
Meta maintains that it is fighting fraud “firmly” because neither users nor legitimate advertisers They want it in the ecosystem, and it maintains internal objectives to reduce substantially This content will be available in key markets throughout 2025.
Even so, internal reviews acknowledge that it is easier advertising scams on Meta platforms rather than Google, and that the company's apps were involved in one third of successful scams in the U.S., according to a May 2025 presentation.
What it means for Spain and Europe
For the European—and Spanish—user, the risk lies in the combination of mass segmentationAI-powered impersonations and a greater margin for disseminating dubious content, which demands examine with a magnifying glass financial, health or retail offers with discounts that are "too good to be true".
For advertisers, this scenario raises costs and erodes trust in the inventory, pushing to demand more transparency on controls, withdrawal thresholds and integrity metrics in each European market.
With regulators stepping up oversight —especially in the UK— and with political pressure mounting, the debate centers on whether the commercial incentives They are aligned with the protection of the user and the legitimate advertising fabric.
The Reuters report depicts a clash between compliance, revenue and reputationWhile Meta defends its progress—withdrawal figures and reduced reporting—internal documents and requirements from authorities in the US and Europe suggest that the challenge remains. structural and will demand stricter controls and faster responses.
