By Ray Birch
ST. PETERSBURG, Fla.—Credit union leaders would be smart to prepare not for a significant disruption to payments as a result of artificial intelligence, but instead for several smaller, less dramatic effects, according to one person.
And credit unions might want to look at what’s happening in Europe and Australia for indications of what might lie ahead, the same person added.
“There's a lot of different predictions right now, and it's always fun to make predictions,” said Brian Scott, PSCU’s chief growth officer. “I've heard a lot of people call this payments 3.0. But I don't think AI is going to make payments 3.0.”
Scott’s comments are appearing here as part of a CUToday.info series examining what new breakthroughs in AI will likely mean for credit unions, financial services and the broader markets.
Earlier reporting in this series on AI includes:
Three Kinds of Data
Scott said there are essentially three kinds of data that create artificial intelligence.
“Descriptive data, so it describes who you are—you are a 47-year-old white male that has five kids, lives in Iowa, those kind of things,” Scott said. “There's predictive data. People are creatures of habit. They shop at the same Starbucks coffee shop every day. They go to the same gas stations. They do their grocery shopping the same day. They spend about the same amount. You can predict Brian's going to go to the same Starbucks every day between 8:00 and 8:30, and he's going to shop at the same Whole Foods every Sunday. We've gotten very good at predicting what's going to happen next.”
In the AI Wheelhouse
It is the third type of data, said Scott, that falls right into AI’s wheelhouse: prescriptive data.
“So, now it’s about being able to put those things together, to prescribe a solution the same way the doctor would prescribe a solution to a health problem,” said Scott. “If you know I shop every Sunday at Whole Foods you can put pieces of descriptive data and predictive data together to create a prescription that’s very descriptive.”
Scott said his spending data show he averages $300 in purchases when he visits Whole Foods. But what happens when he has only $150 in his account?
“Using AI, the issuer could send me a text that I have only $150 in my account. It would create a prescription that says, ‘Before you walk into Whole Foods, do you want to use $150 of emergency cash, or do you want to transfer money’—prescribing a solution to a problem I may not yet know I'm going to have.”
Better Experiences
Scott emphasized he believes AI will help create better experiences for consumers that makes them more satisfied with their card issuers and financial institutions.
AI will also be used in payments, as well, to better thwart fraudsters Scott said. He explained if AI learns consumers’ buying habits, if the technology spots a payment that is well out of someone’s normal habits it can flag that transaction.
“Again, AI helps make better experiences by preventing fraud,” he said.
Avoiding the Drama
Will AI transform payments and will credit unions have to get ready for an entire new word of payments? No, predicts Scott.
“I don't think AI dramatically transforms the whole payments arena,” he said. “I don't think it's massively transformational, like a payments 3.0. I think there's an expectation of, ‘Oh yeah, AI will transform completely how people pay.’ I don't know that AI does that, because I think there's other technologies that do that better. For example, there is starting to be what are called payment facilitators in Europe, and they're making it easy for me to go into a grocery store, for instance, and split the transaction into multiple tenders. I could use Bitcoin for part of it, I could use my credit card for another part of the transaction. I could use reward points. I can make those decisions on the spot and make it easier for me to decide how I want to pay. I think that's more transformational than some of the things AI might do.”
Looking to Other Markets
Where the biggest changes will come with AI and financial institutions, said Scott, is with service and support.
“In Australia many banks have incorporated ChatGPT into how they service and support their customers,” said Scott. “They haven't gotten rid of humans, but they've moved all the transactional type questions—such as what's my balance, tell me about this transaction—to chatbots that can have natural language conversations with people. And it’s no longer that canned-sounding computer voice that is one-dimensional in its conversation. It’s more natural using ChatGPT. They are creating natural language conversations with responses coming back that are very human-like, which is allowing people to have deeper, more important conversations that create better experiences.”
Suggestions Offered
Based on what he sees coming in the area of payments and AI, Scott offered some suggestions for CUs.
“Credit unions should know, and really spend a lot of time understanding, different use cases for AI, especially those that are happening around the world. Because a lot of other countries are way further ahead that the U.S.,” Scott said. “As I said, Australia has banks that have already rolled out ChatGPT into their chatbots. I don't know there's any bank in the U.S. that has done that, yet, to at any scale.”
Scott attributed the slower pace of AI adoption by U.S FIs to how financial services regulators overseas seek and create rules that “enhance creative thinking and technology advancement,” versus the position of U.S. regulators, which he believes are overly restrictive.
‘We Can Learn From That’
“Credit unions should look at what's happening in other parts of the world because they're already more deeply using things like AI, integrating it into their financial institution, integrating it into the payment experience,” Scott pointed out. “We can learn from that, and most of what happens around the rest of the world ends up coming here anyway. For example, open banking and GDPR (General Data Protection Regulation) with data protection, all of those things happened in Europe 10 years ago and they're finally just now coming to the U.S.”
