Spread evenly across a year, the Global Anti-Scam Alliance's estimate of $442 billion in worldwide scam losses comes to about $14,016 every second. It helps to be precise about what that figure is. It describes the pace of losses, not a running tally of individual fraud cases, and it is almost certainly an undercount, because many victims never report what happened to them.
Why a Per-Second Rate Is Useful and Misleading at the Same Time
A per-second rate turns an abstract annual total into something a reader can picture. It also invites a mistake. If you read $14,016 per second as a count of separate scams happening each second, you will misjudge both the size of the problem and the typical loss. A single fraud can take someone's entire savings, or it can take a few dollars from thousands of people. Averages hide that spread.
So the rate is best read as a measure of money moving out of people's hands, not as a measure of how many people are being hit or how badly. Those are different questions, and the data answers them differently.
Three Sources, One Pattern
The surrounding figures point in the same direction, though they measure different things.
The FBI's Internet Crime Complaint Center recorded $20.9 billion in reported US losses in 2025, compared with $3.5 billion in 2019. That is a sixfold rise over six years, and it comes from complaints people actually filed. Reported losses are the visible part of the problem.
The Federal Trade Commission estimates the true cost of fraud to Americans in 2025 at up to $261.5 billion, against $15.9 billion actually reported. The gap between those two numbers is the important part. It suggests that most fraud loss never reaches an official record, whether because victims feel embarrassed, do not know where to report, or assume nothing can be done.
A 42-market survey found that 23 percent of adults had lost money to a scam, and that only 30 percent of those victims recovered any of it. Even where a loss is noticed, recovery is the exception rather than the rule.
What the Gap Between Reported and Actual Losses Tells You
When reported losses are a fraction of estimated losses, the official record understates the problem. Anyone using only complaint data to judge how common scams are will get a distorted picture. This matters for individuals deciding whether a loss is worth reporting, and for anyone trying to understand which scam types are growing.
It also means that a single report carries more weight than it seems. Each filed complaint adds to a dataset that is otherwise thin.
How ScamMarketCap Tracks These Figures
ScamMarketCap keeps these numbers on its data page, and it includes a live counter that starts at zero each January 1 and adds the per-second rate as the year goes on. The counter is a way of showing scale, not a precise accounting of individual incidents. Beside every figure, the page lists its source, so you can check the original report and see exactly what it measures and what it leaves out.
Reading the sources matters more than the headline number. A total labeled "losses" may mean reported losses, estimated losses, or a survey projection, and those carry very different levels of confidence.
What to Take From the Numbers
The honest summary is this: scam losses are large, rising in the official record, and probably larger than anyone has measured directly. The reported figures are the floor, not the ceiling. If you are trying to understand your own exposure or a particular case, start with the scale, then look closely at what each source counted.