The State of Voice Biometrics

Article Featured Image

Voice biometrics—which scores elements of voice calls to authenticate a caller on the other end of the line—is becoming more vital as the need to positively identify callers escalates along with fraudster threats. 

The Year in Review

Last year saw continued adoption of voice biometrics among contact centers, with the penetration hitting 20%, according to Omer Minkara, Aberdeen Group’s vice president and principal analyst for contact centers and customer experience management. “There is a correlation between using voice biometrics and being a best-in-class contact center,” Minkara says. Those that use voice biometrics have a higher customer satisfaction rate—12.9% for biometric users, compared to 5.9% for non-users. Call handling times were also much better for those contact centers using biometrics, according to Minkara, as less time is spent verifying a caller’s identity. 

Financial institutions and others seeking dual-factor authorization also increased their use of the technology in 2018. Unlike PIN and password combinations, voice biometrics uses a combination of 100 physical and behavioral factors, including pronunciation, emphasis, speed of speech, accents, and physical characteristics of one’s vocal tract, mouth, and nasal passages, according to NICE, which offers a voice biometric technology that is text- and language-independent, so a person can talk naturally in whatever language desired. 

Beyond use for verification in the contact center, voice biometrics was also used in 2018 to authenticate people using a larger number of devices, from smartphones to digital assistants. But at the 2018 Black Hat conference, participants reportedly carried out a successful hack of voice biometric technology. According to published reports, the successful breach involved a research team tricking voice recognition by convincing it a machine voice was human. Data scraping and sound editing are other techniques fraudsters attempt to use, though voice biometric providers say their technologies can distinguish between recorded voices and live voices. 

So companies are evolving their voice biometrics solutions as well. For example, in August, NICE launched its Proactive Fraudster Exposure capability based on machine learning technology for its Real-Time Authentication (RTA) voice biometric solution.

The technology is designed to identify previously unknown potential fraudsters by comparing caller behavior with the abnormal behavior associated with fraud in hundreds of thousands of existing call recordings. The suspected malicious actor is then automatically referred for further investigatory steps; if necessary, the culprit is added to a watch list of fraudster voiceprints and blocked from making future calls. The automated process eliminates the expense and time needed for manual checks, the company says.

Similarly, the increased attempts to thwart voice biometrics resulted in a partnership between providers. In December, Interactions—which has its own voice biometrics solution in its intelligent virtual assistant offering—partnered with Next Caller, which uses automatic number identification (ANI) and network information to produce a risk score.

A Look Ahead

In 2019, voice biometrics is expected to continue to grow in usage as banks and others adopt the technology, following its adoption by Swiss retail financial institution PostFinance, Allied Irish Bank, and other large European financial institutions in the second half of 2018. Some of the largest U.S. financial institutions have also started to use the technology, with more expected to this year.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues
Related Articles

Market Spotlight: Security

Voice biometrics is vital, and it's only the beginning. AI and speech technologies are helping drive new levels of security and pioneering new ways to keep the public safe.

Using Voice Biometrics to Improve Customer Experience

Customer experience is at the heart of every brand, but one of the most frustrating experiences for customers often comes when they're contacting you. Voice biometrics may be able to help streamline and improve their experience.

Sensory Releases Embedded AI That Allows Voice Assistants and IoT Devices to Identify a Wide Variety of Home Sounds

TrulySecure Sound ID identifies sounds within the home and can send messages to homeowners.

Embedding Speech Biometrics on Devices

Voice biometrics just might be the next great hope in securing data and other content. But as people use their mobile devices for more and more, embedding speech biometrics on these devices becomes more important