Voice Tech Holds Key to Unlocking Biometric Security’s Full Potential

Speech is powerful — signaling intent, signaling what we need and want.

And speech, in the service of commerce, can be powerful, too.

Collin Davis, chief technology officer of Pindrop, said that we’re a long way from unlocking voice as a conduit to making payments and banking truly hands-free.

He said that value can be unlocked from every voice interaction — whether a consumer is interacting with a financial institution (FI), a digital assistant or even a video chat. Indeed, using voice as a unique identifier can open new capabilities that cannot be unlocked without the proper security that voice verification provides.

To be sure, voice technology is taking on an expanded commercial role. But public mistrust is holding back broader adoption. From accidental orders to deepfake scams, getting voice “right” is a long journey filled with challenges and adversaries.

Davis noted that voice commerce has topped $4 billion and is growing by as much as 300% annually. That growth has been spurred by the rapid adoption of smart speakers, smart assistants and voice search functions built into all manner of devices. For good reason: Voice technology can reduce friction and pain points for companies across a number of verticals — using existing infrastructure.

But even with that growth, as he said, “without a reliable identity capability, it’s very hard to truly rely on voice transactions.” Caution remains the watchword, as roughly 45% of consumers don’t trust the idea of using voice to make transactions.

Firms such as Pindrop help businesses build in the ability to understand more than just what’s being said — using complex voice biometrics and machine-learning algorithms, the tech these firms develop can understand who is behind the spoken words. And they’re certainly garnering attention: Pindrop works with eight of the top 10 largest banks and most of the largest insurers in the U.S., helping them improve their call-center experiences.

And while Pindrop is improving customers’ experiences, it’s also making them more secure.

As Davis told PYMNTS: “If you called into your bank and the bank recognized you right away and said, ‘Hi, it’s great to hear from you,” and they know who you are with a high degree of confidence, they don’t have to go through the security ‘theater’ questions.”

That means doing away with mothers’ maiden names and the last four digits of Social Security numbers — ephemera that’s easy enough to discover on the Dark Web.

Read more: Fighting CNP Fraud With ‘Data Collaboration’

But giving smart speakers and intelligent voice assistants the ability to understand “who” is making requests unlocks their utility for conducting safe transactions while fostering an app-like experience.

Davis drew a parallel between the voice-enabled commerce shift that’s underway today and the advent of advanced biometrics found on smartphones and other devices. It wasn’t until those security features were added to devices that they became tools used seriously in the service of commerce, banking and everyday life.

Closing the Gaps 

To close the trust gap, said Davis, it’s necessary for firms to close the tech gap, too. Application programming interfaces (APIs) allow developers to embed biometrics and voice identity features into apps for mobile devices and other hardware, he explained. And open platforms give consumers more control over how banks actually use their voices — the ability to demand, for instance, that banks erase the voice recordings they use to establish biometric protocols.

Voice is also more efficient than using fingerprints and facial recognition, which have to be verified for each interaction, said Davis. A voice identity can be used and verified in real time across any number of use cases. Among those conceivable use cases: A consumer could walk up to a bank’s ATM and ask it for money, and the withdrawal would be confirmed (or denied, for safety’s sake) based on voice.

“Voice is the way we have been communicating since the day we were born,” Davis told PYMNTS. “It’s the most human way for us to interact with each other. And it is the most human way to interact with machines.”

See also: 55 Percent Of Consumers Would Use Visual Interactive Voice Systems