AU10TIX brings networked approach to synthetic ID fraud detection from payments to all digital identity
A new platform providing a scaling networked approach to combating synthetic identity fraud has been launched by AU10TIX to bring a method used in the payments space to the broader digital identity space.
INSTINCT is an autonomous platform for sharing anonymized identity signals between businesses to fight synthetic identity fraud with a networked approach, internationally-recognized digital identity thought leader and AU10TIX Advisory Board Member David Birch told Biometric Update in an interview.
AU10TIX traditional service consists basically of checking an applicant’s ID document, such as a passport, to determine whether it is real and the data is all true, which produces a yes or no, with FATF Rule 10 applied, which means an account can be opened for that person.
While the principle behind INSTINCT is essentially the same one as AU10TIX has had success with in the financial sector, Birch says, INSTINCT is not just another service in the company’s portfolio, it can be applied broadly to address the problem in the full range of online activities.
“This is driving down a much bigger market, about going into the digital identity space, not just one part of it,” Birch says.
The approach involves generating codes out of identity-related data, and storing the codes for comparison with other identity-related data as it is collected. This allows the network to identify anomalies, and Birch provides the example of a legitimate passport being spotted in use on opposite sides of the world only a couple of hours apart.
“You defend against payment fraud as a network instead of as individual nodes,” he explains.
Adaptive analytics are applied to historical behaviors, emerging patterns and data across the network of participating organizations to detect and share identity risk.
“The use of machine learning techniques in that space has been incredibly successful.” In fact, that success is well-recognized, and Birch himself has been pointing it out for several years.
AU10TIX is a particularly strong position to pull off the transition to networked digital identity protection because of a few advantages the company has, according to Birch.
“You’re taking the tried and tested kind of signals stuff, you’re taking AU10TIX’ incredible data set, because they process so much for all the big platforms, and putting them together in an intelligent way, which is respectful of the privacy and the rules.”
Asked about the role biometrics play in such a system, Birch acknowledges that “the results in that space are pretty good now,” but emphasizes the need for a holistic, continuous and networked approach. Leveraging the network signals approach is how criminal attempts to create multiple accounts with fraudulent credentials that have already been accepted by one organization’s system can be detected and prevented.
Experience with signals sharing and analysis in the payments field, along with the trove of data it has provided, is only part of AU10TIX’ advantage. The other, Birch says, is an architectural decision made early on by the company.
“They opted for a more expensive but fully automated infrastructure, which has been able to scale throughout the virus,” Birch observes.
The heavily-regulated payments space involves enormous compliance costs, and other areas that digital identity is used in come with their own regulatory concerns. For them, manual reviews at scale would be cost-prohibitive.
“We can’t take those costs and impose them on health and education and those other areas,” Birch explains. “We have to start with a more automated approach. It’s taking the automation, machine learning, the signals, taking those into those areas should, I hope, head off some of those costs.”
At the root of INSTINCT’s pitch is something Birch notes in the press release; the stereotype of threats being personified by a misfit individual on his own has become obsolete, replaced by sophisticated organized crime networks. He refers to the fraud problem as a war between networks.
“You need something like INSTINCT to deal with that. You can’t use the old methods. You have to have this kind of infrastructure to deal with that.”
By bringing the approach that has been used to slow the rate of increase in payments fraud to synthetic digital identities broadly, INSTINCT gives organizations a way to protect themselves now, that will become more robust over time as more data yields new insights, and the machine learning is refined.
That evolution is what necessitates bringing the approach used in the payments space, which has long been a target, to other industries that are now relying on online processes as never before. The risks are just as real.
“Synthetic identity fraud once you start talking about health, education and all those other kinds of things, it’s a much more sinister and more pernicious problem,” Birch argues.