Human and artificial intelligence working together to fight payment fraud

The preva­lence of fraud has risen rapid­ly as crim­i­nals have sought to take advan­tage of a unique pan­dem­ic-induced com­bi­na­tion of finan­cial and health threats which have made peo­ple more vul­ner­a­ble to scams. The Nation­al Cyber Secu­ri­ty Cen­tre, the UK’s cyber­se­cu­ri­ty agency, revealed last month that it has tak­en down more scams in the last year than in the pre­vi­ous three years com­bined, fuelled in par­tic­u­lar by coro­n­avirus and NHS-themed fraud attempts.

Though some may asso­ciate the rise of fraud with cyber­crim­i­nals becom­ing more sophis­ti­cat­ed, the real­i­ty is many old scams are among the most preva­lent, with many peo­ple still falling for phish­ing attacks. Younger con­sumers have become well-versed on how to spot such scams, but the growth of dig­i­tal activ­i­ty among old­er gen­er­a­tions, forced to shop online dur­ing the pan­dem­ic, has opened oppor­tu­ni­ties for fraud­sters to tar­get a far less tech-savvy demo­graph­ic. A main entry point to scam con­sumers is when they are mak­ing a pay­ment.

“Many con­sumers are start­ing to make pay­ments online for the first time, going through the process of autho­ris­ing a trans­ac­tion via emails and mes­sages. If they are not very cau­tious, they can be scammed into think­ing sim­i­lar-look­ing emails and text mes­sages look legit­i­mate.” says Tom Pilling, chief risk offi­cer at Trust Pay­ments, a glob­al pay­ments tech­nol­o­gy com­pa­ny. “It is the per­fect envi­ron­ment for fraud­sters to scam peo­ple who are new to pur­chas­ing online.

“We def­i­nite­ly saw a big rise in fraud­sters scam­ming peo­ple that tra­di­tion­al­ly don’t pur­chase online. The modus operan­di of a fraud­ster hasn’t nec­es­sar­i­ly changed in a sig­nif­i­cant way. The meth­ods fraud­sters use to get through fraud engines and fraud trans­ac­tion mon­i­tor­ing solu­tions have cer­tain­ly evolved and improved, but when you look at the trans­ac­tion itself and the infor­ma­tion held with­in it, it is still the case that if it doesn’t look right, it gen­er­al­ly isn’t right.”

The more con­sumers fall for these types of scams, giv­ing away their per­son­al details, the more it also affects busi­ness own­ers and shop own­ers, many of whom have also been forced to embrace ecom­merce dur­ing the pan­dem­ic. Unable to trade from their phys­i­cal stores, mer­chants who pre­vi­ous­ly had no online pres­ence sud­den­ly had to cre­ate a web­site or web shop very quick­ly to sur­vive, while click and col­lect options also grew sig­nif­i­cant­ly.

While the ecom­merce indus­try has raced ahead, pay­ments com­pa­nies have also had to be care­ful to reignite some of their ear­ly edu­ca­tion pro­grammes for mer­chants which were expe­ri­enc­ing sell­ing on dig­i­tal chan­nels for the first time. This has par­tic­u­lar­ly been the case for small­er com­pa­nies, which have need­ed edu­cat­ing on what they should be look­ing out for in terms of typ­i­cal signs of fraud and scams, includ­ing, for instance, card­hold­ers mak­ing repeat pur­chas­es in very short peri­ods of time, aver­age trans­ac­tion val­ue, or veloc­i­ty-type checks.

“These things are quite stan­dard to a lot of big mer­chants that have their own fraud teams in place but they’re new to a lot of small­er mer­chants,” says Pilling. “When you give mer­chants the edu­ca­tion they need, they gain that addi­tion­al lay­er of con­fi­dence. They can always con­sult with us to get advice about par­tic­u­lar trans­ac­tions, but you also see their con­fi­dence grow as they learn to make some of those deci­sions them­selves. If an order looks too good to be true, it should spark cau­tion with mer­chants. That human instinct is just as impor­tant as high-tech fraud tools.”

Mer­chants, and indeed pay­ment proces­sors and acquir­ers, face the dif­fi­cult chal­lenge of not only try­ing to pro­tect con­sumers from fraud but also bal­anc­ing that with the need to pro­vide a strong cus­tomer expe­ri­ence with­out mak­ing the pay­ment process too cum­ber­some. Trust Pay­ments, which pow­ers online pay­ments for some of the world’s most well-estab­lished, as well as emerg­ing, com­pa­nies, has designed intel­li­gent omnichan­nel pay­ment solu­tions that mon­i­tor trans­ac­tions for fraud while help­ing mer­chants grow, by ensur­ing the cus­tomer expe­ri­ence is seam­less and con­ve­nient.

Trust Pay­ments is ded­i­cat­ed to mak­ing as many deci­sions for mer­chants as pos­si­ble with the small­est impact on mer­chant trans­ac­tion autho­ri­sa­tion ver­sus decline. While com­mit­ted to remov­ing false pos­i­tives, it is also focused on ensur­ing all the good trans­ac­tions flow through with­out dis­rup­tion. That means using real-time mon­i­tor­ing to look at dif­fer­ent ver­ti­cals, trends and types of trans­ac­tions to deter­mine what is good and what is bad. Those that sit in the mid­dle, as poten­tial­ly sus­pi­cious, will then be reviewed by a mem­ber of its fraud ana­lyst team.

That bal­ance is cru­cial. Though it’s impor­tant for acquir­ers to be sup­port­ing mer­chants with sophis­ti­cat­ed, machine learn­ing-pow­ered trans­ac­tion mon­i­tor­ing sys­tems, it’s also vital that a human eye is main­tained, both from fraud ana­lysts work­ing for the acquir­er and also from the instincts of mer­chants, who often know the behav­iours of their cus­tomers bet­ter than any­body.

“A lot of it is about learn­ing,” says Pilling. “We con­tin­ue to learn every day with the vol­ume of trans­ac­tions we process online. And with that we’re able to then also look at his­tor­i­cal data to deter­mine how we need to adjust our own rules so that we get that nice bal­ance of good trans­ac­tions, bad trans­ac­tions and those trans­ac­tions that sit in the mid­dle that need fur­ther fol­low up and analy­ses. While our sys­tems are real­ly sophis­ti­cat­ed from that respect, hav­ing that bit in the mid­dle and the human side of things is key to the way we do our own analy­ses and edu­cate mer­chants as to the type of trans­ac­tions that they should be look­ing out for as well.

“There’s absolute­ly no doubt that ecom­merce is a growth area, but we’re also sup­port­ing oth­er emerg­ing ver­ti­cals. As well as our own in-house fraud solu­tions, we part­ner with dif­fer­ent com­pa­nies to pro­vide a unique over­all per­spec­tive of analy­sis. We look at those third par­ties that can do web­site analy­sis, for instance, our cryp­to blockchain analy­sis, to help cre­ate a real­ly unique pic­ture of trans­ac­tions in the future. How­ev­er, we must nev­er for­get that humans have that nat­ur­al instinct, more so than machines, to know if it doesn’t look right, it prob­a­bly isn’t.”

For more infor­ma­tion, vis­it trustpayments.com

Spon­sored by

Trust Pay­ments