Is voice biometrics the key to safer authentication?
Is voice biometrics the key to safer authentication?

It's incredibly important for businesses to have the processes in place to verify customer legitimacy to prevent phone fraud. This is especially pertinent, with research revealing that in the UK, £0.86 per call is now lost to phone fraud—an increase from £0.51 in 2015.  

As identity fraud now accounts for 56 percent of all reported fraud across the country, it's more important than ever that businesses ensure they have the technology to identify legitimate customers and imposters. Combine this with the fact that fraudsters are going to even-greater lengths to deceive call workers and the challenge is on. 

One approach is the development of fraudster profiles; one such includes ‘The Mule', which is identified based on behaviour patterns. ‘The Mule' aims to open a large number of savings accounts – with likely stolen identities – and uses those accounts as vehicles for money laundering and cheque fraud. 

This profile is one of many fraudsters can adopt, and highlights the need for key changes in how companies authenticate. 
Changes in how we authenticate

With traditional face to face authentication predominantly a thing of the past, the industry has moved towards increasingly technological solutions, switching to passwords, PINs and Knowledge Based Authentication Questions (KBAs). These singular modes of authentication have made the process easier for the consumer, but this reliance on technology has left organisations often open to fraudulent activity.

While PINs, passwords and security question answers can be stolen by fraudsters monitoring social media profiles, or through cyber-security hacks, much of our private information like KBA answers can be obtained from fraudsters manipulating their way into accounts. Research from Pindrop reveals that fraud attacks at call centres have increased 113 percent in the past year. 

With a startling amount of fraud occurring over the phone, more businesses are turning to voice biometrics to help them tackle fraud.  

Voice biometrics is the answer

Whilst voice user interface is far from new, the improved reliability has brought it to the forefront of security, particularly when coupled with the news that phone fraud is on the up. 

Last year saw the likes of Barclays launch voice recognition services to protect against fraud on the phone. Additionally, HSBC put a similar service into action for its customers which allows them to securely access their accounts. 

As companies increase their security methods, so do the fraudsters' techniques. If a fraudster knows a company is using voice biometrics, they will attempt to trick the protective barriers. The creation of a multi-layered defence, which covers all forms of authentication from face-to-face, to online and over the telephone, is critical. 

However, for voice biometrics to become a cross-industry security standard, there is a greater need for a more advanced and robust solution. The technology needs to be impervious to background noise as well as the issue that a human's voice alters over time. With the human voice aging every three months, standard voice authentication will quickly become antiquated. This is why businesses must look for a system that can constantly learn and adapt to a user's voice, seeking out any discrepancies which may prove fraudulent.

The next step in protecting your voice

Voice authentication through technology will become commonplace in the near future. As fraudsters continue to become more sophisticated in their means of deception, machine learning-based voice biometrics will soon become the only precaution capable of effectively dealing with the problems at hand. 

The development of solutions for advanced, multi-factor authentication will likely play an important role in the future of voice fraud protection. This technology can be capable of identifying specific elements of each call such as the location of a call, device, whether it's mobile or landline and whether the phone has been used to call the company before.  There are systems available able to analyse 1,380 different elements, helping to detect fraudulent behaviour early, demonstrating the vast potential a voice-based anti-fraud security solution can provide

Contributed by Nick Gaubitch, research director EMEA, Pindrop

*Note: The views expressed in this blog are those of the author and do not necessarily reflect the views of SC Media UK or Haymarket Media.

Nick Gaubitch, Research Director EMEA, Pindrop