The Federal Trade Commission’s January 2022 report put a number on what most people had already noticed: 2021 was a record year for social media scams, with 95,000 victims losing a combined $770 million. The original article on this site covered the 2021 data when it was published. The four years since have been worse — the same FTC tracking shows social media scam losses climbed past $1 billion in 2022 and have not stopped rising. This update keeps the original 2021 numbers as a baseline and adds what changed through 2026, including the AI-generated impersonation surge that defines the current scam landscape.
Contents
The 2021 numbers in context
The FTC’s Januaryreport showed an 18-fold increase in reported social media scam losses from 2017’s $42 million baseline to 2021’s $770 million. About 95,000 consumers filed reports. The 18 to 39 age group was 2.4 times more likely to report being scammed than older age groups — counter to the stereotype that older users are the easier targets. Younger demographics being on social media for more hours per day, and being more willing to transact through it, was the dominant reason.
Each year since has set a new high:
| Year | Reported social-media scam losses (FTC) | Notes |
|---|---|---|
| 2017 | $42 million | Pre-pandemic baseline |
| 2020 | $258 million | Pandemic shopping shift drove the first major spike |
| 2021 | $770 million | The figure this article originally covered |
| 2022 | $1.2 billion | First year past the billion mark |
| 2023 | $1.4 billion | FTC Consumer Sentinel Network annual report |
| 2024–25 | Growth continued; AI-generated scam call-outs first appeared in FTC reports |
The FTC notes that reported losses are widely understood to be a small fraction of actual losses. Most consumer-grade fraud goes unreported — embarrassment, low expectation that reporting will recover the money, and the friction of filing a report all contribute. The true 2021 figure was likely several times $770M.
The three scam categories that dominated 2021
Three categories accounted for roughly 70% of the 2021 reported losses. They remain the dominant categories in 2026, although the methods have evolved.
Investment scams
The largest category by dollar volume in 2021 and every year since. The 2021 wave was dominated by cryptocurrency offers that promised to “double” or “10x” the principal. By 2026 the playbook has expanded: pig-butchering operations (long-form social-media courtship that ends in a fake investment platform), AI-generated celebrity endorsements for non-existent token launches, and increasingly polished fake trading-app screenshots circulated through Telegram and Discord groups.
Online shopping scams
The largest category by number of reports in 2021. The classic version: order an item from an Instagram or Facebook ad, payment goes through, nothing arrives. Meta’s enforcement on dropship-fraud ads has improved since, but the pattern persists, especially around seasonal pushes (Black Friday, Mother’s Day, holiday shopping). TikTok Shop has become a meaningful vector since its US launch.
Romance scams
The third category in 2021 by dollar volume — and the one with the most growth since, because the underlying mechanic is now amplified by AI. About 33% of 2021 victims who lost money cited a virtual relationship as the entry point. In 2026, that virtual relationship is increasingly a chatbot using a real person’s photos, voice, and conversational style. The “long con” timing has compressed from months to weeks.
Why scammers love social media
Social media platforms give scammers four advantages a phone call or email never could:
- Targeting. The same ad-targeting tools that let advertisers find a specific demographic also let scammers find newly-bereaved spouses, recent retirees, or first-time crypto holders.
- Social proof at scale. A fake account with thousands of (purchased) followers and weeks of (AI-generated) content looks credible. The signals users learned to trust on email phishing — a personal account, normal-looking history — are cheap to fake on social.
- Multi-channel reach. A scammer can be in your DMs, on your feed, and tagging you in comments simultaneously. Each touchpoint reinforces the others.
- Friction-free payment. The platforms increasingly host the payment rail (Cash App, Venmo, in-app payments). The hop from “trust this person” to “money has left your account” can be three taps.
What changed since 2021: AI-generated impersonation
The single most significant change in the social-media scam landscape since 2021 is the maturation of generative AI. The 2021 era was still dominated by humans typing real-time. By 2026, the dominant operations use AI to scale:
- Voice cloning — 30 seconds of audio (TikTok video, Instagram story, YouTube comment reply) is enough to produce a convincing voice clone. Scammers use this for the “grandparent scam” (calling an older relative pretending to be a grandchild in trouble) and for fake-CEO calls to corporate finance teams.
- Real-time deepfake video — high-quality real-time face-swap is now consumer-grade. Several large 2024–2025 incidents involved video calls where the scammer wore the face of someone the victim knew.
- LLM-driven chat at scale — a single scammer using a large language model can run hundreds of “romance” conversations simultaneously, each with persistent memory and on-brand responses. The labor cost of long-con romance scams collapsed by an order of magnitude.
- AI-generated celebrity endorsements — fake ads featuring familiar faces (Elon Musk, Mr Beast, Joe Rogan, Taylor Swift) recommending crypto schemes or “free money” giveaways are the dominant investment-scam vector on Meta and X.
The platforms have responded — content credentials, watermarking, AI-content disclosure rules — but the cat-and-mouse moves faster on the scammer side. The single best defense remains user vigilance.
How to protect yourself in 2026
Most of the original 2021 advice still applies. The 2026-specific additions are at the top:
- Verify any urgent money request through a second channel. If a “family member” calls in distress asking for a wire, hang up and call them back on a number you already have. Voice cloning makes this the single most important rule of 2026.
- Treat any video call asking for money as suspicious. Real-time deepfakes are now plausible. Insist on a known-secret challenge (“what street did we live on?”) that cannot be answered by an AI working from public info.
- Avoid wire transfer, cryptocurrency, and gift cards as payment methods to people you have not met in person. These remain the top three payment rails in FTC reports because they’re effectively irreversible.
- Tighten privacy settings. The targeting advantage scammers have starts with the data you publish. Review who can see your friend list, recent check-ins, and life-event posts.
- Verify business profiles before purchasing. Search the company name with the keywords scam or refund. Check the page’s transparency tab on Meta — recently created pages with no real history are red flags.
- Be cautious with virtual friendships that escalate quickly. Romance scams compressed from months to weeks once AI took over the chat workload. A “perfect match” who proposes within four weeks is suspicious by definition.
- Use unique passwords and a passkey or hardware key for any account that has financial implications. See our cybersecurity awareness tips for employees for the full hygiene checklist.
FAQ
Can I recover money lost to a social media scam?
Sometimes, depending on the payment method and how quickly you act. Credit card chargebacks have the highest recovery rate. Bank wire transfers can sometimes be reversed in the first 24 hours if you call your bank immediately. Cryptocurrency transactions are effectively non-reversible once confirmed on-chain. Cash App and Venmo P2P transfers are also generally non-recoverable once accepted. Report to the FTC at reportfraud.ftc.gov regardless — even if you don’t recover money, the report contributes to enforcement.
What is the single most common social media scam in 2026?
By number of reports, online shopping fraud — items advertised, payment taken, nothing delivered. By dollar volume, investment scams — particularly cryptocurrency and pig-butchering schemes. The fastest-growing category by year-over-year report count is AI-impersonation scams (deepfake video calls, voice clones).
Which platform has the most scams?
FTC data has consistently shown Meta’s platforms (Facebook and Instagram) accounting for the majority of reported social-media scam dollars, with Telegram and X following. WhatsApp has grown in scam-vector reports since the platform’s wider business adoption. TikTok Shop is now a meaningful share of online-shopping scam reports specifically.
How can I tell if a video call is a deepfake?
Several real-time tells still work in 2026, though the gap is closing: ask the person to turn their head sideways (most face-swap models degrade at extreme angles), wave a hand in front of their face (occlusion is hard for current models), or have them put a finger to their nose (unusual gestures break tracking). The most reliable test is asking a known-secret question that cannot be answered from public information.
Where do I report a social media scam?
In the US: file with the FTC at reportfraud.ftc.gov, the FBI’s Internet Crime Complaint Center at ic3.gov, and the platform itself (each has a built-in scam report flow). In the UK, report to Action Fraud. In the EU, your national consumer protection agency. Reporting to the platform is what may get the scammer’s account removed; reporting to the FTC/FBI/IC3 contributes to enforcement actions and federal-case patterns.
Share Your Thoughts