To prepare for the AI-enabled future, banks are taking proactive measures to reduce their exposure to potential fraud. Banking leaders are adopting smart approaches to customer identity verification as a way to streamline fraud prevention at the “front door.” The introduction of AI in banking innovations means that customer experience and fraud teams have new tools available to support their efforts, but it also presents fraudsters with a new landscape of opportunities. 

‍

Prove CEO, Rodger Desai, spoke about these challenges on a recent American Banker webinar: “Banks and AI: How to Get a Head Start Using Advanced Digital Identity.” In a wide-ranging conversation with Mike Sisk (Contributing Editor at American Banker) about the operational and strategic aspects of AI and banking technology, and identity verification innovation, Desai emphasized the need for banks to embrace, but be judicious, about AI and other digital banking technology opportunities. 

‍

We have provided the highlights of Rodger’s session below, and hope you will take the time to watch the replay of his American Banker webinar.

‍

The Impact of Generative AI on Fraud Detection and Identity Verification

‍

Generative AI is increasingly getting into the hands of bad actors. Do you think the advances of fraudsters will speed up the need for new identity technology for banks and financial services?

‍

Rodger Desai: Every time there's something groundbreaking and new, there are all sorts of benefits, as well as new risks and dangers. With artificial intelligence (AI) and the algorithms that drive AI models, it makes it easier to deceive people, and we’re already seeing that with misinformation in certain parts of the media. The next step is that fraudsters will find ways to exploit consumers through digital channels. Their machine learning models are getting smarter, and as a result, this is going to be a major issue in terms of how we can protect consumers and fortify organizations. 

‍

With the right type of identity verification, banks are going to be just fine. Consumers get access, they do their business, and they move on. Yet, as banks introduce new types of machine learning and AI, the ability to maintain that level of democratization will become much harder.

‍

What opportunities do you think AI tools present for banks to accelerate digital trust and create more personalized experiences?

Rodger Desai: The one thing that's concerning is how easy it is, with these new tools, for someone to imitate another person. It's not just about how you look, but also how you sound and move – your voice, your movements – all those things can be replicated pretty easily. The question is, how can you create digital keys that can be used to identify a person through those things?

‍

‍

Think of it this way; we all walk around with phones, so it’s possible to use different methods, whether it's SIM cards or FIDO keys, all the various kinds of keys that you can tie a digital identity to. We can use these signals to make passive assessments of who people claim to be. AI can enable banks to use these kinds of behaviors to create better user experiences. They’re also ways for fraudsters to exploit.

‍

A lot of artificial intelligence innovations will actually make it easier to personalize experiences, but banks have to make sure the consumer is who they claim to be. On the flip side, customers have to know that the bank that's offering the service is who they claim to be. But if you can do that and facilitate it in a secure manner, there are great benefits. Banks can expedite customer onboarding and deliver services faster and in more ways.‍

‍

Given the rapid pace of technology development, can you provide some practical steps that banks can take to start preparing for AI and ensuring they have the ability to effectively and accurately validate digital identities?

‍

Rodger Desai: It’s a very difficult task to be able to do that with precision. At Prove, we’ve developed a unique, exact, and accurate way of doing it. If you look at how easy it is to make a phone call where you can go to any country in the world and you can call someone, and there's no friction. It's actually billed, it's a private transaction, typically in a different currency. So, we use those same kinds of approaches when we do our identity authentication. It allows for a passive way to authenticate someone, using an incredible amount of data that gets orchestrated behind the scenes.

‍

Mobile phones are ubiquitous, so using data points from just simple things like the phone number, the length of time a person has had their phone, and datasets and metrics about phone usage enable us to correlate signals to make an accurate decision about the validity of a user. 

‍

Now, there are all sorts of issues that fraud teams have to contend with. These include things like synthetic identities and identity-related regulations. To that last point, one of the core challenges exacerbating this is that a lot of KYC, CIP, and AML processes don't include enough verifications. But we’ve accounted for this at Prove. Consider that, if I’m a fraudster with a stolen phone, I have to have the victim's phone in my hand during the transaction. That just heightens the bar that you have to pass.

‍

How Identity Verification and Fraud Detection Innovation is Improving the Banking Experience for Customers

‍

What banking experiences do you think will be the first to benefit from advances in modern identity verification? 

‍

Rodger Desai: In call centers, some banks are looking at the user just as a “concept” of what customer service looks like; it’s not really addressing what customers need. Increasingly, you're seeing fast-food restaurants where there aren't many folks at all; it's just self-checkout. AI drives the ordering process, and it is a model for what machine learning and AI-driven processes can provide in banking in the future. 

‍

For example, when someone calls a bank's call center or goes to a branch, they're trying to get something done as quickly as possible. If you can predict what they're likely there to do and automate it, that could be a huge operational advantage. Think about some of the repeat tasks that customers engage with service departments for: resetting a password, paying a bill, ordering checks; things that are not terribly complicated. With the datasets provided by AI, you can fill those service needs and even anticipate what customers will want. And so you can see a lot of these tools where there's advanced identity authentication and AI being used to create amazing services.

‍

I’ll give you an example. The other day, I had to wait for 30 minutes to get through to a customer service line. Well, we have the ability now, with AI-driven tools, to predict why people are calling us, what they're trying to get help with, and rapidly deliver what they want. The algorithms driving these results are impressive, and eventually, that will be something we do proactively or before the person calls.‍

‍

Could you share some success stories or best practices from banks or financial institutions that have effectively leveraged these technologies to combat fraudulent activities and enhance customer experiences?

‍

Rodger Desai: A lot of great, well-known brands use Prove as a platform for all identity and customer engagement, and not just as a part of their stack. In other words, every aspect of their account opening processes are driven through the Prove platform.

‍

It's proven to be remarkably successful because Crate & Barrel, Lowe's, and other top brands are getting customers validated with just the customer’s phone number. It’s been set up now so that the consumer not only has to know the phone number, of course, which is easy, but you have to have the phone in your hand during the transaction.

‍

Let’s say that I want to impersonate “Jane Johnson.” I would have to have Jane’s phone in my hand during the transaction versus just knowing things about her such as her name, address, or date of birth. Now, that increases security and privacy but reduces the friction to apply. And because we're not asking who you are – we're just asking you about your phone – it makes it so you can't lie to us, while also making it more private. Because if I don't know who you are, you can't claim to be someone else. And so it has a dual benefit of making it easy to apply with a pre-filled product and being very hard to fool.

‍

At the end of the day, we’re making it easy for the good and difficult for the questionable. And we’re helping companies do that with a pass rate of over 90%. So, whether that's opening an account, signing in, paying a bill, you know, whatever they're trying to do, we’re enabling them to do that without introducing friction and operating expense.

‍

I have another example that comes from our Prove Pre-Fill® identity verification solution, which automates form filling with information we already know – and have validated – about the consumer. Think about when customers are putting in one challenge question. It could be your birth month and day, or the last four of your Social Security number. The ability to ask one question just kind of delineates who's actually behind the action. And in this case, it's on a mobile phone. So you don't even have to put in your phone number because we can derive that from the network and pre-fill the form. 

‍

Considering the new obsession with AI, what does the future of digital identity verification and fraud detection in banking look like, and what exciting developments can we anticipate in the coming years?

‍

Rodger Desai: I think a lot of the things that we're most excited about in the industry, like paying with your palm or just tapping a phone or a card – these are the kind of experiences that people expect now. The challenge, of course, is they're going to be easy to manipulate, replicate, and replay – all the things that these new artificial intelligence tools will allow fraudsters to do. I think it just comes down to basics again, and basically, you always need several factors.

‍

In this case with our pre-filled product, you're tying someone's identity to the perfect key, the SIM card. You know who's doing that transaction, and you can determine whether or not you are holding the phone like you normally hold it. But again, it's tied to the root of trust of that SIM card – holding the phone the way you hold it, or you walking away like you normally walk. Are you doing things that seem like what you would do? But again, tied to the deterministic authentication of a SIM card. Give me a SIM card, and therein are a lot of critical keys.‍

‍

What best practices could you share with our audience today about things you’re excited about for banking technology innovation, and especially ways that identity verification will support these trends?

‍

Rodger Desai: A bank's journey involves coverage and accuracy. If you have an amazing solution but it only covers 1% of the country in terms of consumers, that's not very helpful. If you cover 100%, but the accuracy is only 50%, that's also not very helpful. Everything we do at Prove is about achieving higher coverage and higher accuracy. So, today, I think we cover about 90% of adults, which is amazing coverage. But if you try our pre-fill product at the many places that offer it, you'll see that you and your friends will always get the right identities.

‍

So, I think coverage, accuracy, and, of course, the total cost. As more and more vendors are added, the costs get larger. How can you consolidate a lot of the expense so that you can achieve the coverage and accuracy with the least amount of costs possible? So, those are the factors that we think should form the way to pick your partners, right?

‍

Best Practices for Identity Verification in Today’s Financial Institutions

‍

What role do you see biometrics playing in advanced identity authentication methods, and what will the impact be on the future of digital identity verification for banks?

‍

Rodger Desai: It will certainly help with some practical things, like voice authentication. When my password is my voice, well that's going to be under attack because it's so easy to replicate folks' voices. You can see that online where you can replicate the president's voice, for example. We’re also seeing reports of malware within gaming apps that can track how you hold your phone and how you walk. 

‍

One-Time Passwords, or OTPs, are also important to look at. We’re working with our bank customers to move away from OTPs because we’re seeing so many breaches caused by misuse of them through the use of social engineering. The main thing here is we need to remove our consumers from the equation. Our approach is that we’re not asking who you are until you can't lie, and that makes it easier for legitimate folks. 

‍

What is your perspective on phone verification and authentication versus FIDO?

‍

Rodger Desai: Well, for starters, FIDO is a better approach than OTP. The best thing to do is to start with this: you've got to prefill the application to make it easier for someone to apply. Then you can ask the person, once they're approved if they want to go passwordless, and they can opt-in. And then you now have two keys. You have a FIDO key and an individual key, which is great. And then you can use the FIDO keys and the FIDO key is now on the old phone.

‍

We use FIDO at Prove to see who's dropped off the SIM. You bootstrap back off the SIM again because you're likely to have the same phone number because most people when they change phones, keep the same number. And then you could look at behavioral pieces, like are they holding the phone the right way or they're walking the right way, as they did on the old phone. It'll be more likely they won't be socially engineered. 

‍

Note: The copy above has been edited for brevity and clarity.

‍

‍Watch the American Banker webinar, “Banks and AI: How to Get a Head Start Using Advanced Digital Identity” and learn how customers are using Prove to reduce fraud and improve customer experiences.