Skip to main content

How biometrics can help to reduce fraud and improve the customer experience

(Image credit: Image Credit: Gustavo Frazao / Shutterstock)

Can you tell us how has your technology has helped a company to tackle fraud recently? 

Deutsche Telekom integrated Nuance’s voice biometrics into Deutsche’s existing customer service hotline, which is already powered by Nuance conversational AI-based technology. 

The technology enables customers to speak naturally, instead of navigating a phone menu. Additionally, they can use the sound of their voice to confirm their identity when calling for service. 

Instead of entering cumbersome information such as account numbers, enrolled subscribers simply say the passphrase “Bei der Telekom ist meine Stimme mein Passwort” (which translates as “my voice is my password”). This is matched to their unique voiceprint, akin to a fingerprint, to gain account access. Today already 300.000 customers use the technology and 75% confirm that authentication is much easier. 

How important is the development of voice biometrics for customers and business in relation to fraud prevention?

Even though passwords are largely viewed as outdated, consumers still accept them and businesses still use them. We therefore need to move beyond this archaic thinking that passwords and PINs are the only protection that consumers need.

Instead, enterprises should be protecting their customers through a layer that offers artificial intelligence (AI) based technologies, including voice and behavioural biometrics, intelligent channel, geo and network detectors, as well as conversational and device-focused algorithms. This multi-layer approach provides the highest levels of security since it can identify legitimate customers through the sound of their voice, location, device and the way they talk, tap and type.

By analysing typical conversation patterns, voice characteristics and other aspects of communication, the technology can flag when a call is likely fraudulent and therefore identify offenders whose profiles do not match those of a customer.

To put this into perspective, HSBC recently said voice biometrics has helped it weed out fraudsters – and has prevented over £300m falling into the hands of criminals since it launched in the UK.

Does it have an implication on the customer experience as well?

Today we use hundreds of passwords across our connected lives, and remembering each one can be difficult. Voice biometrics ensures telephone banking is safer, quicker and easier than having to remember often cumbersome passwords.

For example, Barclays Wealth have achieved a score of 90 per cent or higher with 93 per cent of their customers following the deployment of voice biometrics. 97 per cent of clients that have been offered the voice biometrics service have decided to use it as their means of identity verification, leading to a 60 per cent reduction in customer complaints and 60 per cent shorter calls.

What is ‘Pathfinder’ and how can it boost the “conversational intelligence” of virtual assistants?

Project Pathfinder is a breakthrough technology that reads existing chat logs and transcripts of conversations between agents and customers within contact centres, and automatically builds highly effective dialogue models to create and support two-way conversations between virtual assistants and consumers.

Until now, training a virtual assistant to specialise in financial services, telecommunications, healthcare, insurance and other industries was a highly manual, time-consuming, and often lengthy process, prone to human error.

Project Pathfinder demonstrates how machine learning and AI can automate the creation of dialogue models by learning from logs of human conversations. This significantly reduces the time and cost to create an intelligent VA that resolve issues quickly and is capable of a contextual, two-way conversation.

What does that differ from the virtual assistants we have at home? 

Currently the interactions that we have at home with our virtual assistants are still quite limited and can be frustrating – assistants like Siri and Google Home only have certain conversational abilities.

Whilst Project Pathfinder is not yet being incorporated into consumer virtual assistants, it demonstrates that the capability is there for more meaningful conversations with AI – underpinned by existing customer conversations.

How are virtual assistants changing the customer experience, for example within call centres?

To deliver the 24/7 experience consistently across phone, mobile applications, digital assistants or web chat, many organisations are adopting AI-enabled virtual assistants (VAs).

As an example of this, Swedbank Group - the leading financial institution in Sweden, Estonia, Latvia and Lithuania – deployed a VA that uses natural language understanding to understand customer intent and deliver a human-like, conversational responses through interactive, text-based chat.

With 75 per cent of Swedbank customers preferring to conduct their banking via a mobile app or online, the deployment has been a huge hit. The bank saw an average of 30,000 conversations taking place per month within the first three months of adoption and the assistant answering eight out of every 10-customer questions – and is getting better the more as it learns.

Brett Beranek, Vice President & General Manager, Security & Biometrics Line of Business, Nuance Communications
Image Credit: Gustavo Frazao / Shutterstock