The FTC's Romance Scam Numbers Are Worse Than the Headlines Say

Every year the Federal Trade Commission publishes a number that gets quoted and then immediately misused: how much money Americans reported losing to romance scams. The 2024 figure was north of $700 million in reported losses, and the coverage stops there, as if that were the whole story. It isn't. The FTC's underlying data — the Consumer Sentinel Network reports — breaks down who is targeted, how the money moves, and where it starts, and that breakdown is far more useful than the headline total. It's also one of the strongest reasons to look seriously at alternatives to dating apps, not because apps are unsafe by design, but because the open, unvetted contact model is exactly the surface romance scams are built to exploit.

This piece is a close read of that dataset: what the FTC actually measured, what the numbers show once you get past the topline, what they don't prove, and what it should change about how you evaluate where you meet people. It is not a scare piece. It's a look at a specific, well-documented harm and what actually reduces exposure to it.

What the FTC Actually Tracks, and How

The FTC's romance scam statistics come from the Consumer Sentinel Network, a database fed by complaints filed directly with the agency, plus reports forwarded from the Better Business Bureau, state attorneys general, and other partners. Consumers self-report what happened, how much they lost, and how contact began. The FTC then aggregates and publishes annual breakdowns, most visibly through its Data Spotlight series and the Consumer Protection Data Spotlight on romance scams.

That matters for how you read the numbers. This is reported loss, not measured loss — the true figure is almost certainly higher, since many victims never file a complaint, either out of embarrassment or because they don't realize what happened was a scam until much later, if ever. The FTC itself has noted this undercount in its releases. So treat the published total as a floor, not a ceiling.

The Headline Loss Number Isn't the Interesting Part

The dollar figure gets the coverage, but it's a blunt instrument. Total losses are driven heavily by a small number of very large individual thefts — cases where someone was persuaded to invest tens or hundreds of thousands of dollars, often into a fraudulent crypto scheme. Those cases are real and devastating, but they don't describe the typical experience, and citing the total as if it does inflates the perceived randomness of the risk while obscuring the actual mechanism.

The more useful numbers sit one layer down:

• The FTC has reported that reported losses to romance scams have run in the hundreds of millions annually for several consecutive years, with the median individual loss running into the thousands of dollars — not the six-figure outliers that make headlines.
• A large share of contact in these reports starts on social media or messaging platforms rather than dedicated dating apps, though dating platforms remain a meaningful starting point in the data.
• Cryptocurrency has become the single most common payment method cited in reported losses, replacing gift cards and wire transfers as the dominant mechanism — a shift the FTC has flagged specifically because crypto payments are effectively unrecoverable.

That last point is the one worth sitting with. The scam itself — a fabricated relationship used to extract money — is old. What changed is the payment rail, and it changed specifically because it's harder to trace and impossible to claw back. The scam adapted to the tools available; the emotional manipulation underneath it didn't need to.

Where the Real Signal Lives: Age and Isolation

The FTC's breakdowns by age tell a story that doesn't match the "gullible target" stereotype. Reported losses per person tend to rise with age — older adults, particularly those in their 60s and 70s, report far higher median losses than people in their 20s and 30s, even though younger adults file more total reports. Younger people are contacted more often and lose smaller amounts more frequently; older people are contacted less often but the individual losses run dramatically higher.

The mechanism the FTC and researchers point to isn't cognitive decline — it's relationship duration and isolation. Scams that succeed at scale almost always involve weeks or months of sustained, exclusive-feeling contact before any request for money appears. The scammer is patient because patience is what makes the eventual ask feel like a natural extension of an established relationship rather than a stranger's request. People with fewer competing social connections to reality-check the relationship against are structurally more vulnerable to that patience working.

This is the part that gets lost when the story is reduced to a single loss figure. It isn't really a story about naivety. It's a story about how long an unverified, unvetted contact is allowed to run before anyone else in the picture has a chance to notice something's off.

What the Data Does Not Prove

A few things this dataset can't tell you, and it's worth being precise about the limits:

• It doesn't prove dating apps specifically are the primary vector. Social media and direct messaging platforms account for a large and possibly larger share of reported first contact in recent years than dating apps do.
• It doesn't measure prevalence, only reported loss. Self-selected complaint data skews toward people willing to report and comfortable estimating a dollar loss — it says little about near-misses or non-financial harms like emotional manipulation without a monetary ask.
• It doesn't isolate platform design as the causal factor. Verification badges, message limits, and identity checks all vary across platforms, and the FTC's data doesn't cleanly separate which specific design choices correlate with lower reported scam rates.

Where the data is genuinely strong is on the mechanism: sustained unverified contact plus an eventual financial ask, disproportionately routed through irreversible payment methods, disproportionately damaging to people with less independent verification around them. That mechanism is well established across years of FTC reporting, even where the exact platform breakdown shifts year to year.

Why "Verified" Doesn't Close This Gap

It's tempting to read this and conclude that a verified badge on a dating app should solve it. It doesn't, and it's worth being specific about why: most verification systems confirm that a photo matches a live person taking a selfie in that moment. They do not confirm identity, criminal history, marital status, or intent. We've covered this gap in detail in our look at what a verified badge on a dating app actually checks — the short version is that a scammer can pass a liveness check just as easily as anyone else, because the check was never designed to catch them.

Pew Research's survey work on online dating found that a meaningful share of users, women in particular, report being sent unwanted or inappropriate content, and separately, that many users say they've encountered a profile they believed was fake. Those two data points sit next to the FTC's numbers naturally: an open contact model with light identity friction is exactly the environment romance scams are optimized for, badge or no badge.

Where an Agent-Mediated Approach Actually Changes the Exposure

This is the part worth being narrow and honest about. Agent-mediated matchmaking doesn't eliminate the possibility of dishonesty — no system fully can. What it removes is the specific mechanical conditions the FTC's data identifies as enabling: unlimited, unvetted, one-to-one contact initiated at scale by anyone with an account, sustained indefinitely with no outside party ever reviewing the interaction. When introductions are curated by an agent working from a stated set of preferences and constraints, rather than surfaced through an open, swipeable pool, the volume of contact any one person can attempt is naturally bounded, and there's a point of accountability between "stranger" and "date" that a swipe interface simply doesn't have. That's a structural difference in exposure, not a guarantee — the same patience and long-game manipulation that works over a dating app's DMs could in theory work anywhere two people can talk. But the setup that lets it happen at scale, silently, for months, is specifically the open contact model.

This is also the condition under which agent-mediated matching earns its keep: it works best for people who want fewer, more accountable introductions rather than more surface area for contact. It isn't a claim that scams disappear. It's a claim that the mechanism the FTC describes — unlimited unverified access sustained over time — has less room to operate.

What Actually Reduces Your Exposure, Practically

Independent of which platform or method you use, the research points to a short list of things that consistently correlate with lower risk:

1. Treat "we've never met" plus "urgent money need" as a hard stop, regardless of how long the conversation has run. Duration of contact is the scam's main tool, not evidence against it.
2. Move slowly toward a real-world verification point — a video call, a shared connection, a public meeting — before emotional investment outpaces actual information.
3. Never send money via a method your bank can't reverse — cryptocurrency and gift cards are named specifically in FTC guidance because they're the payment types victims can't get back.
4. Keep at least one person outside the relationship informed as it develops. The isolation pattern in the age data suggests this single habit does more work than any platform feature.

Frequently Asked Questions

How much money do people actually lose to romance scams?

The FTC's most recent published data puts reported losses at several hundred million dollars annually in the U.S., with a median individual loss in the low thousands — far lower than the large outlier cases that dominate headlines, but still a real and rising figure.

Are dating apps the main source of romance scams?

Not exclusively. Recent FTC breakdowns show social media and direct messaging platforms account for a large share of reported first contact, alongside dating apps. The common thread isn't the platform brand, it's the open, unverified contact model.

Does a verified badge protect against romance scams?

Not meaningfully. Most verification systems confirm a live photo match, not identity, intent, or history. See our deep dive on what a verified dating app badge actually checks for the mechanics.

Who is most at risk of a romance scam?

FTC data shows reports come from all ages, but per-person losses are highest among older adults, and researchers point to social isolation — not gullibility — as the strongest predictor of a scam succeeding at scale.

Does agent-mediated matchmaking eliminate this risk?

No system eliminates it. What changes is the mechanism: bounded, curated introductions with a point of accountability reduce the open-ended, high-volume contact the FTC's data identifies as the enabling condition, without claiming to make deception impossible.

If you're weighing how you meet people against this specific risk, our overview of alternatives to dating apps and neverswipe both go into the structural differences in more depth.