Voice AI: how it can be used to reduce fraud

Voice AI has the potential to revolutionise how banks reduce fraud and speed up the resolution of fraud claims. There is agile software on the market with integrated voice AI capabilities, creating opportunities that will allow banks to tackle crime, maintain regulatory compliance and improve customer service. While voice AI is still in its early stages, it is an exciting innovation in which banks should consider investing.

Put yourself into the shoes of one of your customers.

You are sitting at home with a cup of tea and have just received a notification via your banking app about a payment made from your joint bank account, which you don’t recognise – a £1,000 flight to Mexico.

You then open the app to find out more about who booked the flight and where it is due to fly from (great that in most countries we can see this info now!). Unfortunately, your partner doesn’t recognise the transaction either, or you suddenly know you’re not going to be able to enjoy your tea.

You start filling out a dispute on your desktop, realise it might take longer than expected, and phone the number on the back of your card.  

Your bank uses voice AI to verify that it is you speaking on the phone. An automated system would take you through a series of additional security questions, without having to any type numbers. It’s a fast and easy process.

That puts you through to a real person who could see you had just been online linked to the dispute. They quickly transfer you to the best colleague to help you resolve the dispute.

The bank cancels your card, sends you a new one, and within 24 hours, the money is credited back into your account.

In this example, the bank has web and app linked to calls and voice and all channels aligned. The good news is with voice AI technology set to advance further over the next three years; the process will be even smoother in future.

The next level will be as follows.

Your bank not only recognises and verifies that it is, indeed, you, who is calling, but also how you are feeling. The AI ‘notices’ tension and worry in your voice as you speak. Using the data from your online dispute form, it calculates the likelihood that you are calling about a fraud attempt on your account. Putting two and two together (using probability-based decisioning), it sends you through to the fraud disputes team direct, should that be the most likely reason for your call. If not, the bank can transfer you to the right person.

Fast fraud resolution via voice AI is a good thing not just for the customer; it is also invaluable for banks. By automating customer identification/verification and using AI to understand the most probable reason for a customer’s call, banks can save huge amounts of time and money.

It also takes pressure off customer service agents, as they can reduce average call duration to mitigate against long call queues. Plus, they can understand the context of each customer’s situation before they’ve even started talking to them. That allows them to start the call on a positive note with the client, cutting the likelihood of complaints, which in turn reduces further work for the team.

What’s more, AI and analytics can be used to keep a record of work done and understand how the bank can prevent other cases of fraud in future to make sure it stays compliant with anti-money laundering regulations.

I remember seeing some of the early voice identification and verification software when running outsourced operations for a bank in India 10 years ago. It was good but limited in application at the time to verification. The technology has jumped ahead so much now that it is possible to fully integrate with an end-to-end customer journey, applying it in multiple steps beyond verification to really improve customer service and stand out from the competition. What’s not to like in something that will pay off for both customers and employees?