Is Preventing Fraud Really Possible in the Era of AI?

Nearly one in five retail transactions occur online, with that figure rising each year. While retailers have adapted their strategies to reflect this consumer behavior, there’s one massive issue that they’re grappling with, especially given the rise in artificial intelligence: e-commerce fraud.

Today’s fraudsters have more opportunity than ever to commit crime, using identity as the gateway to impersonate users, penetrate networks, access sensitive data, and wreak havoc for retailers and their customers. E-commerce fraud is rising in sophistication, speed and authenticity — with users having a difficult time deciphering what's real vs. what's perceived to be the truth. This isn’t slowing down, with nearly half (41 percent) of leaders expecting cybercriminals' use of AI to increase identity threats over the next year.

But is preventing fraud actually possible in this era of AI? Let’s explore some strategies that can help mitigate the threats and get retailers to start embracing AI.

Using AI for Good

AI is a double-edged sword, and retailers need to ensure they’re using the technology for good. Machine learning (ML) and AI can quickly analyze massive amounts of data, using algorithms to identify any unusual patterns in user behaviors. So while manual monitoring is helpful, ensure your team is using AI to catch the needle in the haystack of anomalies, noting this activity and making adjustments over time.

Invisible Risk Signals Help Establish Trust and Extend Sessions

Customer loyalty is won or lost at sign in. Today’s customers crave ease of use, which is what makes the “Keep Me Signed In” option so appealing. In fact, we’ve seen firsthand an increased likelihood of purchase of up to 30 percent from customers that have these long-lasting sessions enabled.

Retailers can also implement threat protection signals that monitor users for risk in real time and extend their sessions based on that continuous risk assessment. This happens without a customer ever noticing via a passive, risk-scoring fraud check that can examine a user’s device, location, IP, behavior or even check for session hijacking. If nothing suspicious is spotted, retailers can reward customers by extending their sessions, ultimately resulting in higher conversion rates. These risk signals help establish trust, giving users a better experience while reducing the cost of fraud.

Password-Less Authentication Over Outdated Passwords

So many passwords can be found on the dark web as a result of thousands of breaches over the past few years. Your password, despite how complex it may be, is no longer enough to prevent fraudsters from logging into your accounts. For this reason, retailers should look to embrace password-less authentication methods, like biometrics vs. the standard status quo password.

Embracing Decentralized Identity

Decentralized identity (DCI) — i.e., providing a digital wallet to customers so they control how they share their personal data — has arrived, with 38 percent of IT leaders having implemented a DCI strategy for fraud prevention this year (up from 13 percent last year).

While government and manufacturing industries have been early adopters, it’s now time for retailers to follow suit. Retailers can take a gradual approach to DCI, starting with applying it to loyalty programs, coupons, and verified discounts via affiliates and multibrand environments. Over time, additional DCI use cases will present themselves. This will also help retailers get ahead of standards like eIDAS2 in the European Union, which mandates support for digital wallets and credentials.

Fraud prevention is possible in today’s AI-powered world, but the onus is on retailers to be more forward-thinking in their approach to identity management. Are you prepared for the challenge?

At Ping Identity, Darryl Jones is vice president, consumer strategy, and leads the Consumer Identity and Access Management (CIAM) solution.