The 2022 State of Voice Biometrics

It’s no secret that cyberattacks designed to gain access to sensitive corporate and personal information have grown not just in frequency but in sophistication, creating an urgent need for stronger defenses that will drive huge market growth for voice biometrics for years to come.

Cybersecurity is a multitrillion-dollar industry that is likely to grow by about 25 percent per year, according to some industry analysts

Fact.MR, for example, estimates the U.S. market for identity theft protection, including voice biometrics, at $10 billion and expects it to grow to $24 billion by 2032, driven largely by government and e-commerce initiatives to increase online security.

Within that space, Research firm Technavio expects the global voice biometrics market to grow by $2.02 billion through 2025, progressing at a compound annual growth rate of 18.81 percent, with the rising prominence of voice biometrics in the banking, financial services, and insurance sectors accounting for much of the increase.

Voice biometrics is one of the principal technologies to which companies are turning to fight fraud.

Technavio noted in its report that accuracy had been a problem with voice biometrics solutions, but the industry has largely come together in just the past few years to push efficiency and accuracy to the point where most user experiences are quick and painless. In many cases, biometric security measures today go largely unnoticed by consumers, though companies typically disclose any use of voice identification at the beginning of calls.

Unlike PINs and passwords, voice is non-transferrable, limiting the ability of voice imitators and recordings to defeat the security. The technology, though, is far from foolproof; deep fakes and other evolving technologies are also growing and threaten to bypass voice biometric security.

Year in Review

“We’ve had another significant leap in the technology,” said Brett Beranek, Nuance Communications’ vice president and general manager for the security and biometrics line of business. “Organizations can authenticate with high levels of accuracy with very little audio.”

Banks and healthcare, two of the most security-minded industries, are adopting voice authentication because it’s accurate and convenient for their customers to use, NICE says on its website.

NICE, which includes a real-time Customer Authentication module as part of its CXone cloud-based contact center platform, says voice biometrics today is easy to use.

To enroll in voice authentication, customers need to call the organization so they can record a voice sample. The voice authentication software then uses the multiple characteristics of the voice to create a voiceprint for the customer. Once the system has a voiceprint, customers will have an enhanced customer service experience, the company said in a statement.

With current technologies, interactive voice response systems can recognize and authenticate customer identities, which enables customers to access self-service tasks, like checking account balances. If the customer chooses to speak to an agent, the authentication information can be passed to the agent; customers don’t need to repeat any information.

Voice biometrics solutions like these are having an impact. A study in the United Kingdom found that losses due to telephone banking fraud fell by 7 percent in the first six months of 2021, with the number of cases falling by 50 percent. The study, by UK Finance, said the decrease was due in large part to the introduction of voice biometrics, which it said is making it harder for criminals to commit fraud.

It’s no surprise, then, that banks and other firms like the technology. The reasons are plentiful.

“Voice biometrics is being used primarily to drive down costs in the contact center,” Beranek says. “Organizations have seen an increase in call volume. Manually authenticating someone by asking security questions or sending someone an SMS message on her phone consumes time. Voice biometrics delivers those efficiencies. The fraud imperative is still there, but now we’re having more balanced conversations about improving operations in the contact center, delivering a better customer experience and simultaneously reducing fraud.”

A year ago, many organizations were still adjusting their contact centers to deal with the pandemic, Beranek explains. “We’ve noticed in the last 12 months that a lot of organizations have stabilized, and we’ve seen a shift from just trying to plug the fraud holes to going back to some of the fundamentals of the contact center industry—delivering quality customer service, reducing average handling times, etc.”

Beranek says that Nuance’s biometric technologies “have grown by leaps and bounds. It’s been the highest-growth solution for us over the last year.”

And it’s not just in the contact center channel. According to Beranek, some organizations are expanding their use of voice biometrics technology into self-service channels. “The use in IVR and mobile apps has expanded tenfold in the last year,” he says.

Voice biometrics is often used as part of two-factor authorization, alongside other vectors like passwords or facial recognition. But while just a few years ago voice biometrics was the secondary factor, today it has become the primary authentication factor.

While proponents of voice verification cite its accuracy and effectiveness, industry experts warn strongly against relying on the technology as the sole factor.

“As deep learning advances, deep spoofing advances as well,” says Todd Mozer, CEO of Sensory. “It’s easy to replicate voices today and have them say any password phrase desired. Real-time detection can help, but passive authentication for voice is not yet perfect.”

Mozer, therefore, recommends layering voice biometrics “with other biometrics that are non-intrusive. We use face and voice together to prevent deep spoofing,” he says.

Voice biometrics has a seamless usage model, Mozer adds. “It works great for devices with microphones—cellphones, smart speakers, TVs, etc.”

Beranek also notes that most people almost always have a voice-enabled device with them, be it a phone, tablet, or laptop. However, video-enabled devices, fingerprint readers, iris scanners, and similar technologies aren’t always as readily available.

A Look Ahead

In its report, Technavio noted that the voice biometrics market has largely been fragmented, and the degree of fragmentation will accelerate through 2025 and beyond.

The report identified Auraya Systems, LumenVox, M2SYS Technology, NICE, Nuance Communications, OneVault, Speech Technology Center, Uniphore Software Systems, Verint, and VoiceIt Technologies as the major market participants.

But they are certainly not alone, and the field is growing quickly. Major technology companies are starting to compete in the voice biometrics markets, a development that Beranek calls a sign of the technology’s maturity and widespread acceptance.

In addition to major tech companies moving into the market, some voice biometric providers have made significant partnership moves recently.

One of the most notable has been an expansion of the partnership between Nuance and Genesys, giving Genesys Cloud CX customers integrated access to Nuance Contact Center AI, which includes biometrics-based authentication and fraud prevention as well as conversational virtual assistants for voice and digital channels.

“The continuing acceleration of digital transformation means providing customers with a combination of advanced capabilities in a CCaaS model,” said Robert Weideman, Nuance’s executive vice president and general manager of enterprise solutions, in a statement. “By integrating our proven enterprise-grade conversational AI, biometric security, and Nuance Mix tooling technologies with Genesys Cloud CX, we are giving organizations the power they need to meet customer expectations today and the flexibility to take advantage of new technologies to address future opportunities.”

ArkX Labs in January started a new partnership to incorporate Sensory’s voice and face biometrics tools within its EveryWord touchless voice solutions. The biometrics collaboration enables ArkX customers to access natural language touchless control features, including support for more than 20 languages, a preset library of wake words, and the ability to create custom wake words. The advanced audio and voice technology supports both human-to-human and human-to-machine speech recognition.

And Pindrop announced a partnership with Five9 to expand its biometric voice authentication, fraud detection, and call intelligence technologies around the world.

“As we head into 2022, a key focus for our business is to make voice authentication technology more widely accessible and beneficial to businesses across a variety of industries,” said Gahn Lane, vice president of global partners and alliances at Pindrop, in a statement.

And going forward, voice biometrics will become more prevalent, working its way down to personal use, Mozer predicts. “Voice biometrics will start to show up with no training required, and suddenly devices will just know who you are and what your preferences are. For example, when I speak to my TV remote and say ‘Play Stranger Things on Netflix,’ why does my TV show a screen where I have to choose who I am? It should just know.” 

Phillip Britt is a freelance writer based in the Chicago area. He can be reached at spenterprises1@comcast.net.