Voice biometrics is one of those technologies: it changes the game. It’s not in wide use anywhere in the world, yet, but it’s going to be the norm for authenticating identity within the next few years. It’s been worked on and refined for almost fifteen years. So it’s at the point where it can be smoothly integrated with an organisation’s other systems, both front and back end. All that remains is to build it into your strategy.
There are compelling reasons to supplement all your other ID and authentication systems with voice biometrics.
For one thing, voice biometrics confers a massive, positive step-change in your organisational security. As with iris and fingerprint recognition technology, it is almost impossible for criminals to replicate the human voice. So, you have a high degree of certainty that the person in contact with your organisation is who he or she claims to be.
Also, voice biometrics has an advantage over iris and fingerprint recognition in that a person’s identity can be verified remotely, using a phone. With iris and fingerprint recognition, the person must be physically present to be authenticated.
In addition, the time involved in verifying someone’s identity can be cut to the few seconds needed for a voice biometrics engine to match someone’s live voice print with one that has been recorded. No more asking for an ID or passport number, a PIN or profile number, addresses, phone numbers, or the name of the person’s dog or first teacher.
Organisations can, as a result, be much more flexible and fluid with their security rules and processes, adapting easily as business strategies and markets require.
For the customer, voice biometrics eliminates the need to remember multiple passwords, usernames, and PINs. Recent research in the United Kingdom should that, on average, people have as many as 100 individual passwords and PINs.
Trouble is, they will often use one or two fairly simple passwords across many different types of authentication, from their bank accounts to their Facebook account. This poses a significant hacking risk not just for the individual but all the companies implicated in that person’s authentication activities. Voice biometrics all but eliminates this risk.
So, why isn’t everyone on the voice biometrics bandwagon?
One of the challenges is that organisations are not clear on the difference between speech recognition technology and voice biometrics. Speech recognition technology is programmed to understand statements or questions from customer and offer a specific response. By contrast, voice biometrics works on the unique tone, resonance, pitch, and biological characteristics of a person’s voice, regardless of the words used. It is the uniqueness of each human voice that makes voice biometrics a security tool.
A bigger challenge to voice biometrics uptake, however, is getting consumers to record and register their voice prints. There are two methods: passive and active. Passive enrolment entails simply recording a person talking and then checking all future calls from him or her against that particular voice print.
However, in some countries, recording someone without their knowledge or agreement is regarded as an abuse of their right to privacy. That aside, passive enrolment is less accurate than the active alternative – and preferred by most organisations.
With active enrolment, people are told what voice biometrics is and how it works before having their participation requested. Their participation is simple and ranges from speaking a sequence of numbers or reading an agreed pass phrase. In future conversations with the organisation, they are asked to speak the same numbers, but in a random order, or to repeat the phrase.
As a rule, when people understand how much time and effort voice biometrics saves them in dealing with the organisation, they participate perfectly willingly. Also as a rule, it’s the early adopter industries, such as cellular operators, that have customers not only willing but eager to use cutting edge technologies. For this reason, a large cellular operators was the first South African company to adopt voice biometrics.
Once the ice is broken, other sectors, such as banking tend to follow very quickly. In fact, the financial services sector has been using voice biometrics for some time to service its high net worth clients. JP Morgan recently authorised a transfer of US$550 million using a voice biometrics password.
In fact, voice biometrics is so effective as an identity check that many governments are considering creating national, centralised voice banks. Citizens and legal residents will then give the organisations with which they interact permission to access their voice recordings.
That’s a few years down the line, though. For now, it’s going to be individual organisations that gain a strategic and operational advantage by using voice biometrics technology.
Integrating a game changer
Because they are game changers, technologies like voice biometrics can’t simply be bolted on to the rest of an organisation’s systems, processes, and strategies. They alter the way an organisation behaves, internally and externally. They hold fairly profound implications for image and operations. For these implications to be beneficial to the organisation across the board, an integrated approach is needed. Not just in terms of systems, but also the impact voice biometrics will have on an organisation’s security value chain as a whole.
It’s vital, therefore, that organisations work closely with specialists in integration of customer interactive solutions, business process optimisation (BPO), contact centre operations, and voice biometrics itself. All of these disciplines affect all of the others. Best, therefore, to understand the inter-relationships upfront and to plan the ways in which you want your organisation to benefit from them.
Darren Arnold, Dimension Data managing executive for Contact Centre Solutions, Middle East and Africa