It’s become impossible to ignore the recent boom in artificial intelligence interest, and while AI is not new to the payments space, its utility is advancing rapidly.
Interviewed for the PYMNTS Executive Insight Series “What Is the Application of AI Innovation in the Payments and Financial Services Space,” Worldline Head of Product and Digital Commerce Gertjan Dewaele said he sees near-term use cases that will make payments smarter and more secure.
“People have seen things like ChatGPT coming up,” he said. “In the years to come, we are going to see a growth of these kinds of AI and AI-enabled technologies across the board. I don’t think payments are any different. Will it radically change how we do things? Perhaps not. But it’ll enable us to do things a lot smarter, with a degree of depth of personalization, of profoundness that we can’t achieve with [non-AI] technologies.”
Those “regular technologies” are rule-based systems that use statistical correlations and less dynamic means, whereas AI is built to learn from each instance and become smarter over time.
Dewaele said he sees this impacting certain areas like fraud detection in a big way. Currently, “there’s a lot of brain power on the side of the payment provider, on the side of the merchant to try to make these rules accurate and based on cases.” AI can do a better job of seeing connections and detailed points between data that humans often can’t spot.
He mentioned the pact Worldline and Microsoft struck in 2021 on using adaptive AI to reduce online fraud, integrating Microsoft Dynamics 365 Fraud Protection into Worldline’s digital commerce payments suite.
Transaction routing is another area where AI can modernize payments, perhaps with even greater effect, as a dynamic AI can find the ideal routing path faster than current systems.
“As a merchant, if I take my transactions that can go to different acquirers or different payment channels, today that optimization is very much guesswork and trying to do a little bit better, seeing what works, adapting,” he said. “Here AI can help, again, make connections between data, but also bring speed to the table that a human configuration cannot achieve.”
This is particularly important for larger merchants sending data to various providers, enabling them to adapt based on authentication rates between providers across multiple payment channels.
Read also: Worldline Acquires SoftPos to Tap Into Contactless Payment Demand
AI also makes a difference in how retailers manage and move inventory at scale by bringing deeper insights into the equation, enabling more conversions.
Pointing to how fast-fashion giant Shein is using AI to detect trends and allow very rapid personalization of their whole supply chain to react, that same processing power can and is being brought to bear to create deeply personalized shopping experiences.
Nodding to action in areas like virtual try-on product recommendations, Dewaele said, “How can I see if this product matches my look? Maybe I want similar products in the catalog of a merchant that could be hundreds of thousands of products. How can I find things that match my shopping behavior and my preferences? [Those are] problems where AI can create an interesting user experience.”
While merchants and financial services firms considering AI may think it’s the tech integration that’s difficult, it’s more a case that their data needs to be readied for ingestion by AI systems.
“The bigger challenge for a lot of businesses trying to reap the advantages of AI is not necessarily how to install an AI technology in their software landscape, but to get the data ready and this whole data pipeline set up to bring all this information from all their scattered systems into whatever is that AI,” he said. “It’s the same in payments.”
See also: Worldline and ING Expand Card Partnership
Drilling down on AI’s present and future in fraud-fighting across payments and financial services, Dewaele said “a big part of the tradeoff in fraud detection is to what extent do [merchants] want to put rules and limitations and things like a 3D secure check or an authentication while still having a good customer experience? On that side, AI can help as well.”
He detailed how the Worldline Labs team is innovating with device-level behavioral biometrics in the gaming space to the point where an AI-enabled system can detect if an unauthorized user is trying to make in-game purchases using someone else’s controller.
Back to business use cases, many companies are using AI to secure and improve back-office functions like accounts receivable (AR) and accounts payable (AP). While Dewaele said he sees the application, he conceded that getting AI to work well in these cases will require more effort, but ultimately it remains the same thing: finding hidden connections in big amounts of data.
An example of this is aligning recurring payments with paydays, which differ between countries. He said data mining reveals that certain days of the month show spikes in failed payments. Delving into that, Worldline data scientists saw the correlation to paydays. AI can help here too.
“As a company that’s doing subscriptions, figuring that out and looking smartly at your data can help you maybe by just putting the renewal date for your subscriptions in that country a few days later,” he said. “Again, it boosts auth rates and has [fewer] canceled transactions.”
While less convinced about things like AI-controlled autonomous payments in the near term, Dewaele said, “I wouldn’t want an AI to automatically order my groceries for me … but I would love to have a grocer that when I log into their website to make an online order say based on all your previous purchases … ‘We’ve already [assembled] half of the order for you.’”
We’re always on the lookout for opportunities to partner with innovators and disruptors.
Learn More