sp3n - stock.adobe.com
Fair Isaac Corporation, the credit scoring giant commonly known as FICO, revealed two new AI-driven products aimed at protecting users' digital identities.
FICO Identity Proofing enables customers to onboard new users by letting them take selfies holding a government-issued ID, while the FICO Authentication Suite introduces new authentication controls, including biometric, multifactor and behavioral analysis controls.
The FICO AI products are powered by EZMCOM, vendor of a digital security, risk and fraud platform that FICO acquired in August 2019.
"The acquisition of EZMCOM adds the AI suite of tools to their product portfolio, which definitely adds an edge," said Moutusi Sau, senior director and analyst at Gartner.
While FICO has long sold AI-driven software for detecting and preventing financial fraud, it hadn't sold products that specifically combat digital identity fraud, said Liz Lasher, vice president of fraud and financial crimes at FICO.
"When it comes to security, one of the biggest gaps is in the idea of the digital identity," Lasher said. "If you have consumers that don't have an easy way to identify themselves, then there is going to be fraud."
FICO Identity Proofing helps confirm consumers' identities by enabling them to send a selfie alongside a photo of a government-issued ID. Machine learning models then analyze identifying factors on the ID, including the holographic mark, and a deep learning model runs a facial recognition match between the ID and the selfie, Lasher explained.
Moutusi SauSenior director and analyst at Gartner
Also, models help determine if a selfie is taken in real time by evaluating, while the camera is running, whether a person is acting like a real person, such as performing small actions such as blinking and fidgeting.
The other new FICO AI product, the FICO Authentication Suite, helps automatically protect consumers' digital identities using multifactor, biometric and behavioral authentication, such as keystroke analysis and device telemetry, that can run client side or server side.
With keystroke analysis, machine learning models detect patterns in how a user types and inputs information, and flags instances in which a user deviates from that pattern.
"If you think of how you type or how you enter your password, that's actually a pattern signature. Only you type that way," Lasher said.
Device telemetry involves using machine learning models to track device-related patterns, such as preferred language setting, preferred browser and device IDs. Again, if the model detects a deviation in the pattern, it will flag the inconsistency. That factors into a risk-based score that determines the probability of fraud.
Such authentication tools are a basic requirement for financial services organizations, Sau said.
"Most of the use cases like 'know your customer' and 'anti-money laundering' require biometrics and behavioral analysis," she said. "A lot of fraud detection tools have behavioral analysis."
The new FICO AI digital identity products are not unique to the market. Other companies such as Mastercard's Brighterion also have fraud prevention products on the market.
"Brighterion is the strongest in terms of fraud detection amongst the three," Sau said. "But for ID Verification, FICO has made a strong acquisition."
FICO differs from the competition by selling a connected platform that includes FICO Identity Proofing and the FICO Authentication Suite, as well as products for dealing with financial fraud, credit risk, compliance and customer engagement, according to Lasher.
Experian and DataVisor
A day before FICO's announcement, Experian, the major credit reporting agency, detailed a new partnership with DataVisor, an AI-driven fraud prevention and cybersecurity company.
DataVisor offers products that use supervised and unsupervised machine learning for fraud detection and prevention. The acquisition gives Experian access to new fraud analytics capabilities, which FICO already has, according to Sau.