The spoken word is something that we very much take for granted. We typically start at about a year of age, and continue to do so for the next seventy years or so. The words themselves however only convey a part of the message – there are infinite nuances to voice communication based on situation, mood audience and physiology. In fact, an individual’s voice is as unique as a fingerprint – no two are the same.
This attribute is often overlooked. The value of voice is rarely assessed and certainly, from a telephony perspective, voice is as commoditized as tapwater. However, the ability to use voice as a biometric signature may be coming of age, particularly as we are on the cusp of a whole host of mobile transaction services where authentication of the end user will be paramount.
There are four factors of authentication - something you have, something you know, something you are and something you do. If you go to a supermarket and pay with a credit or debit card, the transaction uses two obvious factors – something you have (the card), and something you know (your PIN) or do (your signature). However, there is a more subtle factor at play – your physical self. You have to present the card in person and interact with the clerk. At their discretion, they could decide to decline the transaction based on your behavior, your body language and yes, even your tone of voice.
In the Anywhere world, the physical component of being at the point of purchase is removed, meaning an increasing requirement for validating that you are who you say you are and that the merchant is also who they claim to be. Online card fraud is a case in point – often only one factor of authentication is present – the card details. The merchant is essentially flying blind, having to trust that you are who you claim to be. Consequently, card fraud is rife, with merchants subject not just to losses, but higher fees for taking “card not present” transactions. Reciprocally, cardholders have been victims of fake merchants sites in “phishing” attacks, where card details are harvested from unsuspecting consumers and then used to commit card fraud using their credentials.
Given the ubiquity of mobile devices, global connectivity and the naivety of end users, fraud over mobile is set to grow in parallel with the anticipated explosion of m-commerce and m-payments transactions unless robust authentication can be introduced. Retrofitting the 4+ billion mobile subscribers with biometric capture technologies such as fingerprint and retina scanners is clearly out of the question. Passwords and PINs could be used, but as has been proven so many times, these are far from robust and easily compromised. So, the obvious candidate for the job is voice biometric technology. It can work with any existing device, is highly robust and, moreover, uses something that we are intimately familiar with - there is no learning curve to speaking into a phone.
For mobile operators, this presents an opportunity to rejuvenate voice revenues, with value being added as an extra layer as voice becomes more than just speaking but a means of uniquely identifying an individual. Even in the Anywhere world, voice could become what it has always been in face to face situations – proof that you are who you say you are.
SOURCE