Meta Earns Billions from Fraudulent Advertising

Last year alone, Meta is said to have earned around 16 billion US dollars thanks to fraudulent advertising. Meta disputes the figures that Reuters is reporting on. But even if the numbers are lower, we all know about plenty of scams and AI garbage on Facebook and Instagram that we can see and experience ourselves every day. It seems astonishing that Meta's automated systems cannot (or do not want to) recognize much of this as fraudulent attempts.

Reuters refers to internal documents, which were allegedly leaked to them. These are forecasts for the year 2024. In them, Meta estimates that the various platforms would serve people a total of around 15 billion high-risk ads per day. High-risk means that they are concrete fraud attempts, i.e., links to websites with fraudulent intentions, to illegal gambling sites, and to prohibited products that are sold—for example, in the health sector.

A real masterstroke by Meta: if there is suspicion of fraud, the ad is simply madepricier at first. Only when the probability of fraud is over 95 percent do the systems react and block the ad.

How Meta Combats Fraud and Scams

Meta says Reuters is presenting a “selective view” and distorting the company's approach to fraud. While a spokesperson does not dispute the figure from the report at hand that Meta estimated to earn around ten percent of its revenue from fraud, he also says the estimate was only rough. Actual earnings were significantly lower. “We aggressively combat fraud and scams because users of our platforms do not want such content, legitimate advertisers do not want it, and we do not want it either,” says the spokesperson. This is also already stated similarly in the report that Reuters has.

However, it is repeatedly noticeable that the systems apparently do not function particularly well. At least, many people have the feeling that Facebook now consists almost entirely of made-up headlines, dubious ads, and suggestions that one does not understand why one is receiving them. Nevertheless, it must also be said that the fraudsters are getting better and better. It is certainly often a kind of cat-and-mouse game. Fraudulent ads are now often hidden behind legitimate-looking, supposed articles. Websites and matching images are becoming easier, more realistic, and faster to create thanks to AI. It is also noticeable that fraudulent ads appear much more frequently on Facebook than, for example, on Instagram. This may be due to the higher average age of the users.

In this regard, the Meta spokesperson also tells Reuters that the number of reports of fraudulent ads by users has decreased by 58 percent in the past 18 months. In the year 2025, Meta had already removed 134 million fraudulent ad contents. However, this figure is not offset by any statement about the total number of ads.

Reuters also reports that potential penalties would be lower than the revenue from fraudulent ads. Although there are new regulatory steps in the USA, Great Britain, and with the Digital Services Act (DSA) also in the EU, according to which Meta could be increasingly held responsible for the content on the platforms. But so far, there have not even been any penalties, but only investigations have been initiated.

(emw)