Meta allows impersonation scams to run rampant as political deepfakes hawk phony benefits

By now, it is hardly a secret that social media, and Facebook in particular, has become a garbage dump for bad information. Already targeted by foreign actors as fertile ground in which to sow discontent and partisan hostility, its inextricable ties to AI development mean it is now awash in so-called AI slop: memes, images and advertisements created with generative learning models.

It’s an ideal habitat for scammers. A new investigation by the Tech Transparency Project (TTP) has found that scammers are spending heavily on Facebook ads that use deepfake videos of prominent political figures to push bogus government benefits. They come from groups with names like I Love My Freedom and End the Wokeness, and Get Covered Today.

“Scammers who use deepfake videos of Donald Trump, Elon Musk, and lawmakers like Alexandria Ocasio-Cortez and Bernie Sanders to promote fictitious government benefits are among the top political ad spenders on Meta platforms,” it says.

Sixty-three 63 scam advertisers are responsible for 150,000 political ads on Meta platforms, spending a whopping 49 million dollars – putting them among the top political ad spenders on Facebook and Instagram in 2025.

Scams target seniors with dubious incentives

The bait is fake stimulus checks, government spending cards and Medicare payments, alongside old fashioned investment scams and snake oil sales. (“Get a free ICE hat just by paying this mandatory packaging fee!”) The quarry is seniors, who are young enough to be on social media, but often lack the digital literacy needed to spot scams and deepfake media.

“The findings show how scammers are taking advantage of advances in artificial intelligence technology, public confusion around the status of social safety net programs, and lax Meta content moderation to target new victims,” TTP says. “Meta is allowing this activity even though it prohibits scams and says it invests in scam prevention to keep users safe.”

“All of the 63 scam advertisers showed they had ads removed by Meta within the past 12 months for violating the company’s policies, though nearly half (30) continued to advertise as of Sept. 30, 2025.”

TPP says Meta did not respond to its request for comment, but did make a statement to the New York Times, which reported on the findings, saying it will “invest in building new technical defenses.”

Time to stop believing what Meta says

Social media companies have been promising to make their platforms safe and secure for as long as there’s been doubt about their safety and security. Yet the fraud problem keeps getting worse: according to the Pew Research Center, nearly three quarters of U.S. adults have experienced some kind of online scam or attack. Some older adults have lost their life savings to scammers.

It is past time to stop taking what Meta and its offshoots say at face value. The firm says it is “constantly working” to weed out scams. Yet a May 2025 report from the Wall Street Journal asserts that “current and former employees say Meta is reluctant to add impediments for ad-buying clients who drove a 22 percent increase in its advertising business last year to over $160 billion.” Meta knows some of these are scammers – but the money’s too good to get rid of them.

TTP’s research, conducted using Meta’s Ad Library, found that Meta subjects political ads to an “enhanced level of scrutiny.” Nonetheless, it found dozens of advertisers running scam ads targeting U.S. users: “for example, during the 90-day period from April 3 to July 1, 2025, at least 45 scam advertisers spent more than 18 million dollars on Meta platforms.”

In its conclusion, TTP says that, “even when Meta disables these accounts or removes the ads, the damage has already been done to consumers – and the company has already profited.” It believes that the mass availability of generative AI tools is likely to further turbo-charge scams “unless Meta takes steps to improve enforcement of its own policies.”

This seems unlikely to happen in any comprehensive way. However, there are efforts being made on the biometrics front to try and stem the bleeding. Meta already has a facial recognition system that flags suspicious celebrity faces in ads (so-called “celeb-bait”) and matches them to the profile photo of the celebrity in question, to determine if they are real. Last month, the government of Singapore ordered Meta to deploy its biometric systems to crack down on government official impersonation scams.

Regardless, the current status underlines a concerning truth that is becoming more evident by the day. To quote a statement issued by U.S. Senator Bernie Sanders, whose likeness has been used to push fake benefit scams: “AI is advancing rapidly, but the laws and protections we have in place for everyday Americans are dangerously outdated.”

Stronger regulations for AI are presumably not high on the list of priorities for a U.S. administration that, right out of the gate, tried to block any AI regulation at all for ten years. As such, the problem is likely to get worse before it gets better.

Article Topics

AI fraud  |  biometrics  |  deepfake detection  |  deepfakes  |  Facebook  |  facial recognition  |  generative AI  |  Meta  |  social media