It’s a scary fact, but anti-fraud experts agree that cybertheft rings are now well-funded multinationals with revenue goals of their own. Even worse, they now have our personally identifiable information (PII) and are using it to open fake accounts, decimating online trust.
Compounded by the loss of in-person cues with the rise of eCommerce along with the sunsetting of browser cookies, traditional knowledge-based authentication (KBA) is becoming less effective as fraud rings accumulate and deploy stolen data, often successfully.
For these reasons, NeuroID CEO Jack Alton told PYMNTS’ Karen Webster, “I think 2022 and 2023 will be the years of behavior. I think the overall market is understanding that simply relying upon third-party data sources that are historical and fact-based are not going to be adequate. Adding behavior to the mix, especially upfront, top of funnel, can really help us get more from those investments that we’ve made today and deliver on much better experiences.”
With customer acquisition costs (CAC) rising and convincing frauds responsible for opening millions of accounts, businesses are between a rock and a hard place, turning away good customers who are incorrectly flagged while letting fraudsters waltz right in.
Monitoring behavior during account opening and transactions to detect fraud is one way that highly sophisticated fraudsters and their bot armies can be stopped. As Alton said, “It’s looking for new ways, better ways to figure out how to assess, confirm or screen identity at scale, because the traditional means of clicking the submit button, going out to all my third-party data sources, the fraudsters are getting all those answers right.”
When cybercrooks either possess or can easily buy your PII on the dark web, the better way to ferret out fakes is watching how they pretend to be you, at scale. That’s what Alton calls “the crowd view,” and it’s very revealing during the prescreen that NeuroID performs.
“One of the really exciting things about behavioral as a technology is that it’s very difficult to fake if you are who you say you are from a behavioral perspective, and it can’t be compromised,” he said.
Calling it “an elegant solution to add another layer of visibility that’s been missing,” Alton said long-term memory and one’s own digital footprint are “your possession. That’s impossible to compromise, whereas the facts can be compromised.”
That’s why top-of-funnel identity screening, at the crowd level as well as at the applicant level, is stepping into the breach.
See also: Legacy Fraud Techniques No Match For Cybercriminals Using ‘Weaponized’ Identity Credentials
Three Telltale Signs
Regardless of how advanced their own tech and methods are, one thing even deft fraudsters are bad at is acting like the person or entity whose PII they’re using. That’s detectable.
Alton told Webster that the NeuroID platform looks for three telltale fraud signals.
Number one: “Is this a human or is this a machine? Behavior really gives powerful indications into that,” he said.
“The second thing we’re looking for is, yes, this is a human, but are they using their own identity, or are they using a list of stolen identities that they’ve purchased from the web? The third one, which is really the biggest problem we have today, is the person who they say they are, they’ve interacted with their PII as though it were theirs, and we want to remove friction for that person. That is the applicant level screen that we’re doing at scale.”
That 1-2-3 approach to frictionless upfront identity screening “allows you to use the platform that you’ve built and the tools that you’ve accumulated much more effectively.”
It turns out that even the most gifted fraudster encounters great difficulty in trying to act like the identity fraud victim when attempting to use their credentials and PII.
Tipoffs involve entering PII out of order, referring to documents to answer questions the actual person would know by heart, entering names in incorrect order, giving the wrong address, or taking much longer than the average customer to complete the onboarding process.
Anti-fraud measures analyzing behavior must also be on the watch for bots, he said. “When we look at automated behavior, like bot behavior, they don’t make mistakes, they move through the application at the exact same pace, they typically move through much faster than a human could move through.”
By shining a light onto those behaviors, behavioral screening assures identities, “helping our customers and our platforms see that this is who they say they are, this is not who they say they are, they’ve compromised the information, or this is a machine that’s doing this at scale.”
NeuroID’s Crowd Alert solution does this, observing the behavior of large groups opening accounts or transacting, with artificial intelligence (AI) and machine learning watching millions of interactions for those that stand out as unlike the others for all the wrong reasons.
For this reason, Alton said, “Our Crowd Alert product is really hitting a nerve in the market. If you think about the sophistication of fraud rings and bot attacks post-pandemic, now that the federal funding and other things that were low-hanging fruit have dried up, they didn’t just go out of business. They are essentially businesses of their own. They’re funded now. They did really well. They stole about $200 billion last year. They have growth projections that they’re trying to hit. What they’re trying to do is, they’re trying to automate it.”
With one large customer reporting that fraud attacks coming from bot rings are up 350% year over year in 2022, Alton said, “What the markets really needs is proactive alerting to changes in the crowd, and then isolating that subpopulation of higher risk so that they can zero in.”
“It allows them to apply the resources and the technology they have to decide if they want to do business with those people. But the more important thing is it keeps the door open for the good customers, so that they’re not being lumped in as a fraud or bot attack.”
See also: AI Helps Businesses Tell Friend from Foe
Combatting Fraud and False Declines
With CACs doubling in many cases and businesses turning away good customers mistakenly while letting fraudsters in, the crowd-level view is becoming essential for online businesses.
Segmenting applicants into genuine and neutral or risky groupings has key advantages, Alton said. “One, you have to stop the bleeding on these sophisticated fraud ring and bot attacks. But two, if you’re not creating a fast-track path for your genuine customers, somebody’s going to. With the cost increase for filling top of funnel, that’s probably more important now than ever.”
NeuroID users are finding immense value in being able to view “subpopulations of people that that are familiar with the information that they’re inputting versus unfamiliar, or if you look at a subpopulation that wasn’t a human, what’s compelling is just how dramatically different they are. It’s not nuanced. It is very different,” he said.
With fraud at historic levels and companies getting more restrictive in their onboarding processes as a result, Alton believes crowd-level behavioral analytics are now an “existential” consideration as fraud steals with one hand and closes the new business aperture with the other with each hit.
Noting that companies clamp down onboarding after each subsequent fraud event, he said, it’s “narrowing the customer set that can come through. It’s really exacerbating the problem. We believe that technology like ours is no longer just for the innovative companies. If you keep narrowing the size of your funnel and putting your best customers through additional friction, you’re going to eventually find a point where that will not sustain your business.”
Adding insult to injury is that the good customers being turned away are not a small number, with Alton saying, “We have firms now that are measuring the size false declines [and] misidentifying my good customer as 70 times larger than the global fraud problem.”
It’s an opportunity for NeuroID because, as Alton told Webster, “My sources say that the volume is going up, not down. That just means there’s an absence of a better way of doing it. That’s what we’re spending all day, every day on doing, is bringing a more frictionless opportunity to identity screening and verification.”