Meta: A scam machine? Reuters uncovers billion-dollar scam involving fraudulent advertising at the tech giant

An analysis of internal documents, which are said to date from 2021 to 2025, suggest that Meta failed to effectively curb the flood of fraudulent ads for at least three years. According to Reuters, the social media company allegedly displays around 15 billion daily ads to users that show clear signs of fraud. These high-risk ads alone allegedly generate approximately $7 billion in revenue for Meta annually.

Calculated risk: Penalties are cheaper than abstention

The documents reportedly indicate that Meta deliberately takes a hesitant approach to combating fraud. If its automated systems detect fraud with less than 95% probability, Meta doesn't penalize suspected scammers with a ban, but rather with higher advertising rates (penalty bids).

A particularly serious allegation is that, according to Reuters' research, Meta allegedly engaged in ruthless internal cost calculations. Potential regulatory fines were estimated at up to $1 billion, significantly less than the "revenue" from the high-risk scam ads. Meta is expected to generate approximately $3.5 billion annually from these ads. According to the Reuters report, there is a strong suspicion that Meta is deliberately basing its approach on the assumption that fines are cheaper than foregoing fraudulent revenue.

Priority: Artificial Intelligence (AI), at the expense of security

Reuters further reports in its exclusive story that the high revenues from fraudulent advertising (scams) are also being used to finance expensive AI projects by Meta CEO Mark Zuckerberg. A capital investment of up to $72 billion is planned, according to the report. At the same time, Meta is said to have strictly capped the revenue losses for its security teams: In the first half of 2025, the Meta fraud prevention team was not allowed to take any measures that would cost Meta more than 0.15% of total revenue (approximately $135 million).

Users are the ones who suffer. According to Reuters, the consequences of this prioritization are reflected in user data. In 2023, Meta ignored or incorrectly rejected approximately 96% of the roughly 100,000 valid fraud reports from users each week. Reuters also reported that the stark example of a hacked Air Force recruiter, through whose account a colleague apparently lost $28,000 in crypto fraud, illustrates Meta's negligent standards to the detriment of its victims.

Meta contests the allegations

A spokesperson for Meta, Andy Stone, denied the allegations. He said the documents presented a "selective view" and distorted the company's approach. The internal estimate of 10.1% of revenue coming from prohibited advertising was "rough" and too high, as it also included "many" legitimate ads. Stone highlighted that Meta aggressively combats fraud and has reduced user reports of fraudulent ads worldwide by 58%.