Social media platforms, especially Facebook, have become a hotbed for payment scams, putting financial institutions in a position of playing the last line of defense to stop money transfers from a victim to a criminal.
The Federal Trade Commission tallied the toll on consumers last month in a new report: $2.1 billion in losses in 2025 for consumers who said they were victimized by a scam that started on social media. Consumers reported losing more money to scams that originated on Facebook than on any other social media platform, the report said.
A Consumer Federation of America report this year cited similar earlier findings. Among online platforms, 57% of reported scams were on Facebook, according to CFA’s March report, which cited 2024 Better Business Bureau data.
“Facebook is the worst in terms of safety for people from scams,” said Ben Winters, director of AI and privacy at the Consumer Federation of America.
Winters said he suspects this is the case because Facebook is one of the longest established platforms and many people have accounts. Facebook also has “a really wide variety of attack vectors,” including direct messages, comments, advertising, pages and Marketplace.
The other top two sites for scams cited by the CFA’s report were Instagram and WhatsApp, which are also owned by Meta Platforms, Facebook’s parent company.
Those harms are also prompting lawsuits.
This month, a county government in California sued Meta in a landmark civil case, alleging that Facebook and Instagram profit from scam ads while consumers are defrauded. That follows another lawsuit by the organization Tech Justice Law against Meta last month on behalf of CFA making similar allegations.
For Meta’s part, the company is investing in tools and technology to combat payment scams. Meta is deploying AI to analyze text and images to detect impersonations. It created Facebook alerts if a user receives a friend request from an account that shows signs of suspicious activity and rolled out warnings on Marketplace related to common scam patterns, such as when a seller asks for advanced payment through Zelle.
It has set up a Scam Prevention Hub with consumer education resources. In the next few years, the social media giant plans to roll out advanced AI systems to more accurately find scams and remove impersonated celebrities.
Meta, however, can’t stop scams alone. Fraudsters may use Facebook to groom targets, but the fund transfers occur off platform, according to a Meta spokesperson, who said true deterrence requires a collaborative effort with governments and financial institutions.
“Scams are growing in scale and complexity and we are aggressively combating scams across our platforms with advanced detection systems,” the Meta spokesperson said.
Fraud shows up on Facebook
Payment scams on social media platforms, including Facebook, take myriad forms.
- Fake investments. Paid ads or private groups lure victims to send funds into false investment platforms or crypto schemes. This type of scam resulted in $46.6 billion in losses in 2024, according to CFA’s report.
- Romance or friendship schemes. Also referred to as “pig butchering,” a criminal exchanges several messages over time with a victim, grooming them and aiming to build trust. Once trust is established, the fraudster steers the victim to make deposits or investments.
- Facebook Marketplace scams. Bad actors list items for sale below market prices and then pressure them to move the transaction off platform to apps such as Whatsapp, said Kevin Lee, field chief technology officer of trust and safety at Sift, a fraud prevention platform. The buyer pays, but the item never arrives and the seller’s account disappears.
- Impersonation fraud. Criminals create fake pages impersonating celebrities, making their authority and schemes seem legitimate to victims.
AI scales up scams
The problem is only set to worsen as artificial intelligence tools enable criminals to cast a wider net, automate scams and more realistically persuade their targets. The FBI’s Internet Crime Complaint Center received more than 22,000 complaints last year reporting AI in scams and fraud, amounting to about $893 million in losses.
With AI, “in the time that [it] used to take to generate one fake listing, you can now create 10,” Lee said. He added that AI can correct spelling or poor grammar if a non-native English speaker originates the scam, which makes the communication appear more professional.
AI enables fraudsters to build fake websites, clone voices or create deepfakes with another person’s face, said David Maimon, a professor at Georgia State University in Atlanta and head of fraud insights at SentiLink. In his research, Maimon has tested agentic tools and seen them develop an entire code for a scam, including creating fake social media profiles and having automated conversations with hundreds or thousands of people.
“Agentic AI will simply bring it to a different scale,” he said.
Whose job is payment scam prevention?
Financial institutions are taking steps to educate consumers and prevent scams. JPMorgan Chase, the largest U.S. bank, introduced a feature last year to limit Zelle payments originating from social media. A Chase spokesperson said about half of Zelle scam claims reported by its customers start on social sites.
The challenge, however, is that “the payment itself is often the final step,” said Justin duPont, head of Zelle product at JPMorgan Chase.
Scam payments can be difficult for banks to spot because the user initiates the transaction, making it seem legitimate, Maimon said. Once a victim is at the point of making a payment, it’s often too late for banks and payment firms to intervene.
“Now the victims are simply willing to give the money away,” Maimon said. “It’s very difficult to break the spell.”
For that reason, financial institutions and consumer advocates are pushing for action upstream, saying the social media platforms have a shared liability to combat scams. Maimon said social media sites are certainly making an effort, but “we need to see more of it,” particularly in flagging potential fake profiles.
Meta is “doing less than most other platforms” to limit scam ads, Winters contended. In fact, the company reportedly generated $16 billion from scam ads, Winters said, citing an investigation by the media outlet Reuters in November.
During an April 15 House Committee on Financial Services hearing on combating fraud, Rep. Brad Sherman (D-CA) called on social media sites “to do a better job of policing the ads that power their companies.”
Meta removed more than 159 million scam ads on its platforms last year, the company said.
Lee, who served as a global spam operations manager for Instagram and Facebook from 2014 to 2016, said scam prevention was “absolutely” a priority. He focused on identifying and blocking false accounts that impersonated people in schemes for financial gain.
While Lee isn’t privy to Meta’s internal efforts today, he sees social media companies taking proactive steps, particularly using AI to identify suspicious accounts or bot-generated content. “As much as AI is contributing to this challenge, it can also be part of the solution,” Lee said.
