Make better decisions on the fraud prevention market

Fraud prevention is hot right now, and the biggest names in payments are paying up for it.

We've found a lot of interesting signals. For example, in October 2025, Feedzai raised $75M at a $2B valuation, after being picked to help the European Central Bank fight fraud on the digital euro. And in August 2026, Visa agreed to buy BioCatch, which spots fraud from the way people type and swipe, for $2.4B in cash. As AI makes scams cheaper to run, banks and payment networks are paying more for the tools that stop them.

In this section, we look at why fraud losses keep rising despite better tools, which kinds of fraud are growing fastest, and what these deals mean for independent fraud startups.

Every fraud team faces the same trade-off: stop more bad payments without turning away good customers.

We compare six kinds of fraud prevention companies, from payment fraud and account takeover to new-account fraud, e-commerce fraud, chargebacks and fraud decision platforms. Their buyers range from banks and card issuers to online merchants, marketplaces and gig platforms. For each, we rank the pains that hurt most, like scams where victims send the money themselves, false declines that cost sales and fake accounts made with AI.

This shows which fraud losses a company will pay to cut first, and which ones it still writes off as a cost of doing business.

Fees per transaction or per check, yearly platform licenses, chargeback guarantees and shares of recovered losses: fraud prevention companies charge in very different ways.

We compare who pays and where the costs sit. A per-transaction fee grows with volume, but big merchants push the price down. A chargeback guarantee wins deals fast, but the vendor pays when its model gets it wrong. Bank platforms earn large licenses after long sales cycles. And every vendor pays for data, from device signals to identity records.

You'll see which models keep their margins as fraud gets smarter, and which ones get squeezed when payment providers bundle fraud tools for free.

Fraud keeps growing, but that alone doesn't explain why companies spend more to stop it. Here, we go through the real drivers.

Instant payments, more online shopping and AI tools that make scams cheap to run are the drivers everyone sees. Several others push in the same direction. Every fraud attempt a network sees makes its models better for every customer. Rules that make banks repay victims give them a reason to spend before the loss. And once a company sees how much fraud it was missing, it tends to add more checks.

Before building or investing, you should know what makes a company pay to stop fraud before it happens, rather than just absorbing the losses.

Forter and Riskified come up first when merchants talk about fraud, but they cover only one corner of the market.

Much of the action happens away from the checkout page. Some startups only stop scams where victims are tricked into sending money. Others spot fake accounts on gig and dating apps, catch money mules at banks, or fight false chargeback claims for merchants.

Side by side, they're tough to judge: each one sees a different kind of fraud, and companies rarely publish how much they catch. A vendor that blocks 99% of bad payments may also block too many good ones, and that number almost never makes the pitch deck.

Below, you'll find the market grids and ecosystem maps we've built to show which kinds of fraud each company stops, and where new players are getting in.

Sardine, Alloy, Signifyd… a few fraud companies have turned stopping fraud into steady revenue.

Sardine combines device and behavior signals to stop fraud at banks, fintechs and crypto firms. Alloy lets banks and fintechs check new customers and watch their accounts on one platform, and reported more than 900 customers in July 2026. Signifyd reviews online orders and repays merchants for fraud on orders it approved. Each picked a clear kind of buyer, though none has yet shown it can cover every kind of fraud for every company.

We'll look at how these teams pooled data from many customers, priced their risk so they could survive a bad month, and widened their product from one kind of fraud to the next.