Facebook uses a combination of AI and human moderators to flag advertisements. But the company’s human moderation is “entirely inadequate,” and it’s not clear how many scams its AI flags before they reach users, said Paul Bischoff, the editor of Comparitech, a site that rates security software and has monitored illegal Facebook ads
“We don’t really know how big the problem is,” he said, “but there’s obviously still a lot of them getting through.”
The ads reviewed by The Markup are unlikely to have met the company’s standards for ads. For one, Meta’s rules include tight restrictions around any cryptocurrency ads. Potential sellers must meet specific eligibility requirements
, then submit a form to Facebook for approval before they begin to sell ads.
Advertisers on the platform also must be careful about how they associate themselves with Facebook. Ads may mention “Facebook”
so long as it’s not the “most prominent feature” of an ad. Using the company’s corporate logo is prohibited, and ads cannot imply an endorsement
. The company’s policy doesn’t specifically mention use of “Meta.”
Some of the pages serving the ads were removed before The Markup reached out to Meta for comment, and the company removed others after The Markup’s request for comment.
“The ads flagged to us violated our policies against deceptive and scammy behavior so we removed them,” Meta spokesperson Mark Ranneberger said in an emailed statement. “Our systems get better when people report this kind of behavior in ads by tapping the three dots in the top right corner and selecting ‘Report Ad.’ ”
The ads aren’t the only example of Facebook dealing with imitators on its platform. In 2018, The New York Times reported
on how fake Mark Zuckerbergs were scamming Facebook users, enticing some with a fraudulent “Facebook lottery” win and then requesting payments before receiving the cash. The Times uncovered hundreds of accounts on Facebook and Instagram impersonating Zuckerberg and Sandberg.
Media personalities in multiple countries have filed suit
against Facebook after their images appeared in cryptocurrency scams, and in 2019 a court in the Netherlands ordered the company to more proactively stop scam ads that feature celebrity images.
In a report released last year
, the Federal Trade Commission said reports of cryptocurrency fraud had “skyrocketed” and that almost 7,000 people had reported a total of more than $80 million in losses between October 2020 and May 2021—an increase of 12 times in frequency and 1,000 percent in money lost, according to the agency.
Around the internet, imitation is a strategy that pays off for those scams.
Twitter, for example, has spent years dealing with scammers on its platform attempting to borrow the identity of Elon Musk.
Some savvy hackers have, in the past, taken over
verified Twitter users’ accounts, switched the accounts’ profile photos to Musk’s image, and claimed to offer massive cryptocurrency rewards in exchange for a relatively small investment of cryptocurrency. In its recent report on cryptocurrency scams, the FTC said it had received reports of Musk impersonators taking more than $2 million in just six months.
According to the agency’s report, people in their 20s and 30s “reported losing far more money on investment scams than on any other type of fraud, and more than half of their reported investment scam losses were in cryptocurrency,” while users 50 and older were relatively unlikely to report being victims of such scams.
The ad Facebook served to a Markup reporter offering a chance to get in on the ground floor of “META Token” was targeted toward American men between the ages of 30 and 64 and offered them a chance to “be one of the first buyers” of the new currency.