Nuance shares voice biometrics fraud prevention successes in EAB talk

Voice biometrics avoid some of the common challenges around authentication for financial institutions, and have resulted in huge savings through reduced fraud for Nuance customers, as attendees heard during the European Association for Biometrics’ (EAB’s) latest virtual lunch talk.

Nuance’s Victor Gomis, a fraud and biometrics executive, presented ‘How to prevent fraud using AI-biometrics,’ which focussed on user authentication for fraud prevention.

Nuance recently became part of Microsoft, which Gomis says presents new growth opportunities for the company.

Common fraud approaches have changed over the past decade, shifting from large numbers of credit card-cloning operations to more industrialized approaches, leading to data breaches, amid an overall 126 percent increase in records compromised, according to the Identity Theft Resource Center.

The dark web facilitates sales of personal records to identity theft rings.

Gomis shared data on synthetic ID fraud, and reviewed the legacy methods banks and other organizations have employed to protect themselves and their customers. These methods, like one-time passwords and device-printing, have high false positive rates and deliver poor experiences for users. He advocates not for these methods to be abandoned, but rather combined with biometrics, in line with PSD2 recommendations to employ multiple authentication factors.

Businesses seem to be generally aware of the value of biometrics for authentication, with Cisco estimating a 48 percent increase in adoption from 2020 to 2021, Gomis notes.

On-device biometrics, however, do not provide the identity assurance needed for strong authentication, however.

In addition to modalities like behavioral and voice biometrics, Gomis suggests that conversational interactions can yield authentication data, such as flagging the language patterns behind social engineering attacks. Those three modalities are included in Nuance Gatekeeper.

Voice and financial transactions

Gomis explained the basic concepts behind voice biometrics, starting with a physiological explanation of speech. The advantages of the biometric include transparency and easy use.

PSD2’s multi-factor authentication requirement has been addressed by many retailers in ways that increase friction, harming conversion rates, Gomis points out. Voice is an attractive option for those seeking a better way, he says.

He also presented data from a Nuance client that showed attack rates falling over time once a voice biometric watchlist was implemented. NatWest reports a 300 percent return-on-investment from Nuance’s voice biometrics in the first year, according to the presentation.

Nuance clients have found fraudsters using the voice channel and the digital channel together to carry out fraud attacks, such as by carrying out social engineering on a call agent and then utilizing the information in an app.

Gomis also demonstrated successful authentication and fraud detection with Nuance Gatekeeper for the EAB audience.

Article Topics

authentication  |  biometrics  |  biometrics research  |  EAB  |  European Association for Biometrics  |  fraud prevention  |  Nuance Communications  |  synthetic voice  |  voice biometrics