AI Isn’t New to Payments but Use Cases Like Predictive Forecasting Are

The current explosion of interest in artificial intelligence (AI) may seem new to the uninitiated, but truth be told, this cutting-edge tech has been silently performing tedious, high value tasks for years.

Whether it’s in payments, retail or numerous adjacent areas, Andrew Stucchio, vice president of global pricing and analytics at Discover Global Network, tells PYMNTS the power of AI in the financial services space still has a long way to go.

“We’ve been working through and leveraging AI and data for many years across Discover Global Network in a variety of capacities, on our fraud detection and monitoring side, in our merchant data and classification side,” Stucchio said, laying out some of its prime use cases.

Fraud detection and protection are areas of financial services where AI-powered solutions have been working in the background for several years. “We see that as an area where there will constantly be challenges as fraud emerges in new and present ways across our space,” he said.

“AI enables us to optimize customer experiences, but all while minimizing fraud and compliance issues,” he said. “We see it with anti-money laundering, account takeovers, bot testing, and even for synthetic fraud and collusive merchants.” Discover Global Network is leveraging AI as a distant early warning failsafe for potential fraud events across its network and for its partners.

“That’s probably a big space where I see its biggest relevance for us.”

See also: Discover on Leaving Legacy Payments Tech Behind in 2023

AI as Change Agent

Asked about the perception — and reception — of AI among businesses using Discover Global Network, Stucchio said it depends on the company and the vertical, although many want in at this point.

Outside of but adjacent to financial services, he noted, “I think we can all agree that over the last three years, retailers have had quite a bit of a struggle to meet supply chain challenges that have been brought upon by the pandemic and some more recent geopolitical disruptions. I see many retailers embracing AI because it helps them” better manage these issues.

He pointed to two emerging areas where AI has the potential to be a change agent: predictive analytics and various forms of automation, be it in the warehouse or the retail store.

“With predictive forecasting, I would say that many firms are using AI so that they can be more accurate in determining demand, down to some of the most sub-regional levels possible. That’s helping them optimize their inventory,” he said.

As it pertains to automation, Stucchio sees “retailers embracing AI because it enables them to use digital workers, robots and play out the ability to enable repetitive tasks to happen automatically and without error. You can see that in the case of some of the document processing happening in the back office. You can see that with some of the transport automation that is happening for the logistics side of things. You see it with quality control.”

These applications go along with the fraud mitigation and anti-money-laundering (AML) use cases where AI models have been toiling away for the past several years, looking for needles in transactional haystacks with a level of scrutiny humans can’t match at scale.

He said that fact “really helps spur some of the willingness of many partners across many industries to increasingly embrace AI for all of those advantages it serves.”

Read: Discover Says Debit’s Resilience Drives New Payments Opportunities

Payments Customization Smartens Up

With a portfolio that includes DiscoverDiners Club International®, and PULSE®, Discover® Global Network already has AI actively working on different aspects, from security to payments specialization.

“In our space, we’re using AI for many instances, and one that comes most to my mind right now is our Discover Commerce exchange, or as we abbreviate it, DCX,” he said. “This is our scalable data and commerce platform that offers us a 360-degree view of clean, enhanced merchant information, and it does that down to the transaction level.”

DCX supports financial institutions by tailoring credit and debit cards to work in specific ways, guaranteeing “that their cards work where they want, when they want, and in the way that consumers want to pay.”

An example of how AI is applied in this way, Stucchio said, “as I think about DCX in the context of AI, AI is powering what we call at Discover a restricted authorization network, which in essence is enabling these white-labeled cards powered by our Discover Network to be configured the way that the issuer or the program wants them to be configured.”

That can mean using AI to customize cards for use at specific merchants, for example, by implementing spending protections and limits for minors. “With AI through our DCX program, in essence, what it’s doing is giving us a reliable way to identify and classify merchants even when the data itself is quite messy,” he said.

Even while the world, and the financial services sector, alternately marvel at and profit from the use of AI, it’s not a set-it-and-forget-it proposition.

Stucchio told PYMNTS that “as you think about more sophisticated models that leverage AI or machine learning, I think one of the big watchouts for us is that while they provide greater accuracy, they also have a tendency to degrade at a faster clip than some of the more traditional models that might be out there.”

For that reason, among others, he thinks utilizing AI in finance requires “the right level of governance, risk management, and monitoring of models to ensure that they sustain that level of predictability. That was the whole reason and sort of justification for them in the first place. That’s a word of caution I heed as I think about AI and as we leverage it here at Discover.”