AI has had to run fast to catch up with dramatically different patterns of behaviour during the pandemic. David Bailey, Santander’s Head of Payments Strategy, and Andy Renshaw, VP for Banking Solutions at anti-fraud software firm Feedzai, consider the short-term and long-term consequences of changes in the way we spend.
Artificial intelligence (AI) software spots bank account fraud by identifying behaviour changes. A sudden interest in iTunes from an account held by an 81-year-old that has never even featured an online purchase before would be an obvious red flag. But what happens when seemingly everyone switches from buying clothes and holidays online to face masks and toilet rolls? That was the scenario that faced bank security systems from February onwards as COVID-19 swept across continents.
With big increases in home-delivered groceries and a collapse in restaurant and travel costs, almost everyone’s buying patterns changed as countries locked down. A report by the MIT Technology Review in the US examined how the top 10 Amazon search terms were knocked off the chart and replaced by sanitisers and protective wear, reflecting anxiety around the pandemic. And some AI programs built to act upon subtle shifts were thrown into chaos as a result.
David Bailey, Santander’s head of payments strategy, says the crisis underlines that AI still needs human help to make sense of the bigger picture and intervene when needed. “We use machine learning and AI, particularly in financial crime and fraud, to help us identify patterns of behaviour that we would deem suspicious,” he says. “But it’s always been complementary to the existing tools that we use, and we’ve always had to combine it with some human interaction – human intuition – to guide its direction.”
For Andy Renshaw, vice president for banking solutions at fraud software firm Feedzai, changes as the pandemic began were manageable because, overall, the behaviour was still recognisable. Feedzai screens bank accounts and debit cards in real time, using hundreds of data points to pinpoint anomalies. The immense volumes of data it handles require several data centres to run concurrently, but Renshaw says the systems held up well.
“We’ve had a more stable machine learning and AI performance than we originally feared, right from the start. “The MIT Technology Review asked the question ‘is AI cracking under the pressure of the coronavirus pandemic?’,” he says. “Fundamentally, machine learning models look at individual behaviour and, from our perspective, it involves huge volumes. But, although the coronavirus pandemic is unique in terms of the scale of change, the underlying principles behind it are not unusual.
“Machine learning software is already used to spikes on paydays, Black Friday, Cyber Monday, the last shopping day before Christmas, etc. With coronavirus, although customers are behaving in a slightly different way, their transactions are still relatively normal.
“They’re still using merchants that they used before, for high-value and low-value payments. Fraudsters, too, are still, in many ways, committing the same types of fraud, and those transactions still look unusual, relative to normal behaviour. “So, although the mix within ‘normal’’ has changed, our systems have remained stable,” says Renshaw.
A desired increase in contactless payments as shops deter cash use to minimise the risk of transmitting the virus has been one of the success stories of the crisis, with the transaction ceiling rising from £30 to £45 within days in the UK and by similar amounts elsewhere. But, depressingly, fraud levels have risen too, as criminals targeted consumers who were new to digital payments and online shopping – meaning banks have had to re-examine their balance between security and low friction.
For Santander, the increase in criminality has meant the bank has had to continually remind customers to stay alert to the dangers.
Bailey says: “There were all sorts of scams that suddenly appeared, it was astonishing how quickly fraudsters picked up on the opportunities. But, on a positive note, we’ve seen late adopters to contactless take off with the pandemic. The industry should be proud that it could quickly get together and push the limit up to £45, and
I think that made customers relax a bit.” Renshaw agrees that fraudsters have sought to capitalise on the crisis, but says the frauds are not new, they have just been given a COVID-19 spin.
“The fraudsters are leaping on the weaknesses and exposure points that COVID has brought,” he says. “Also, though we’re not seeing it yet, an economic downturn brings specific frauds to the fore. So, things like hardship fraud, refund chargebacks and claims for non-receipt of goods will emerge.” On the subject of friction in payments, both men recognise the need for balance. And balance is not just vital from a security standpoint, it’s needed to ensure bank customers feel comfortable. Bailey says: “Many people expect there to be a level of friction in the journeys they see because they want their finances to feel secure. They’re happy to remove the friction when it’s a very trusted brand, and in those instances leave their card on file with them. But if it’s a new provider, people are hesitant.
“Legislation is pushing us towards more security but, from an open banking perspective, there are opportunities for less friction, especially for the big digital platforms where people trust those brands.” Renshaw has experience of an app launch where users were worried by the fast speed of payment processing and actually requested proof that security checks were being made. Reassurance, he says, was established by a message, essentially telling customers to ‘please wait a few seconds as we conduct checks’. But the security advances made by AI mean the general trajectory is towards less delay and fewer clicks in the customer journey, not more. And, with a nudge from retailers, bank customers can be guided through major changes, believes Bailey, who says ‘change will be bigger than people expect, but slower than they expect’.
Consider the Amazon Go model, where scanners register what customers take and charge their account in the background as they leave. There’s real potential for that model to appear in the future, but I think it’ll happen quite slowly,” he says. “By way of personal example, I’ve started using filling station apps when I fill my car, so I don’t need to go into the shop to pay. But I still feel compelled to wave my phone at the attendant before leaving, because I feel uncomfortable with just driving away! I know they get a notification, but I still want to wave. So, I think it’ll take time for consumers to get over the thought of just walking out without making a physical transaction.
“Payment journeys, or payments as part of broader journeys, will become far more integrated in a way that feels seamless to the customer. Micropayments for media consumption may become a new model to reduce the reliance on advertising, for instance – it will be interesting to see what changes happen but it’ll be bigger and slower than we expect.” Renshaw believes changes in payments will grow from the bottom – cash use will continue to fall, while lower value contactless and micropayments increase.
“But it’s different at the moment when you go up to higher value payments, or transactions where you expect to have a long-term relationship with the vendor. Maybe you’re buying a car and you’re going to be returning to that dealership on an annual basis. Is that something you’re comfortable doing via a contact-free payment? I’m not sure people are yet. Another trend is the wallet, or membership model, where consumers register their payment details with a provider they trust. That will continue to grow.” Trust is the cornerstone, but the sudden growth of contactless payments to reduce risk during the pandemic is a perfect example of how people are willing to embrace change.
“There are customers who haven’t been digital or ecommerce users with their supermarkets, for example, moving to being digital users,” says Renshaw. “Probably what we’ve seen is customers looking at where they can establish trust. People can then progress to not just using a different channel, but to interacting with companies that they haven’t used before.”