AI promises 2025 advances for payments industry
Artificial intelligence has already changed the payments landscape in ways the average consumer and business owner may not notice, but in 2025, the emerging technology might make itself more visible as we pay for goods and services, according to analysts and academics who follow the payments industry.
Fighting fraud, for example, is one of the key practical applications of AI. But it is something a consumer is unlikely to think about unless their credit card number is stolen or their bank account is hacked. As the technology advances, however, consumers may find that it gives them more payment options and more suggestions about what to buy.
While programs like the chatbot ChatGPT have made headlines in recent years for imitating humans in some limited conversations and occasionally writing convincing essays, the programs that most financial technology and payments companies use are older than that, stressed John Wilson, director of the MS Fintech program at the University of Connecticut’s School of Business.
“The techniques that many people are calling AI, they’re not innovative, they’re not new, but they’re happening perhaps at a larger scale because of computing power,” he said.
That computing power, however, means that artificial intelligence may change the way people pay in several key ways.
The consumer side
AI programs could result in more options at checkout terminals — including buy now, pay later — by making it easier to offer those options.
Gen Z likes to pay quickly and have multiple ways to settle a transaction, said Davi Strazza, president of North America for the Dutch financial technology platform Adyen, citing findings from the company’s own survey. Artificial intelligence can help merchants that want to offer those options by analyzing a consumer’s transaction history and instantly creating a risk profile so that consumers can immediately use alternative payment methods like BNPL, he said.
“There’s a new generation of shoppers that care a lot about what those terminals offer,” Strazza said.
Fintech companies are also going to use AI to browse a customer’s transaction history and needle them to make more purchases, said Michael Imerman, an assistant professor of finance for the UC Irvine’s Paul Merage School of Business and author of the book ‘The Economics of Fintech.’
“A lot of the data that’s generated from payments, especially when combined with behavioral analytics, can be used to get insights into customer preferences, and then recommendations can be made to the customer,” he said.
Companies like Amazon and PayPal already do this, but artificial intelligence will turbocharge the process, making it much more ubiquitous, Imerman said.
“It almost seems like they’re reading your mind, but it’s really just taking the digital footprint that you created from your transactions, and being able to put content in front of the customer that is in line with their preferences and their interests,” he said.
Unexpected card declines will also grow less frequent partly due to artificial intelligence, said Bill Maurer, director of the UC Irvine’s Institute for Money, Technology and Financial Inclusion.
AI can instantaneously create a risk profile for a customer, based on data that is not typically included in a credit report, such as their spending and payment history, he said.
Artificial intelligence can immediately take into account “things like the regularity of bill payment or the regularity of a remittance,” Maurer said. “It can see that every three months I’m sending $100 to my mother in Mexico. That’s not going to show up in a traditional credit report, but something like that can give a bank or credit union a degree of information about my credit risk.”
The merchant side
On the merchant side, artificial intelligence has the potential to reduce friction and increase conversion rates in the payment process.
TriumphPay, a Dallas-based payments platform that caters to the transportation sector, will use AI to better process invoices.
The company serves as a go-between for companies that sometimes process invoices in different ways, and can leverage artificial intelligence to bridge those differences and settle transactions faster, said TriumphPay President Melissa Forman-Barenblit.
“Each of those parties will be able to validate that an invoice is real and then execute a payment,” she said.
AI can also help spot irregularities and mistakes, Forman-Barenblit said.
“If there’s an issue or missing document or some required piece of information that’s not on that documentation, we use AI to be able to identify those exceptions and be able to quickly get that information,” she said.
The emerging technology will also automate more tasks to smooth out the payments process behind the scenes, said Rocio Wu, a principal at F-Prime Capital, a Cambridge, Massachusetts-based venture capital firm that has invested in payments companies such as Toast.
“There’s so much administrative burden and human labor that’s going to the back-office to either reconcile payments and close the books every month or just process payments, especially for a business that has a large volume of transactions,” she said. “It’s a huge burden.”
Artificial intelligence is becoming advanced enough that it can start to take some of that work from human employees, Wu said.
For instance, AI can help assess the risk associated with a new customer, said Ravi Adusumilli, executive general manager for the Americas at Airwallex, a Singapore-based financial technology company.
“You can deploy different AI agents and models to be able to get information that’s out there to complement what the customer is telling us,” he said. “Then you can have a more educated response on ‘is this customer going to be a high risk or low risk?’”
Preventing fraud and scams
The speed with which artificial intelligence can sift through increasing volumes of data makes the prevention of fraud and scams one of the most practical applications of AI, and the role of the emerging technology in that fight continues to evolve.
The payments software company ACI Worldwide has been using AI for years, but it’s applying the increasingly powerful tool in new ways. At the moment, the company uses AI to help its customer representatives be more efficient; to assist its software developers; and to detect fraud at scale.
“You can find these patterns much quicker and more effectively using AI,” Tom Warsop, CEO of the Elkhorn, Nebraska-based payments technology company ACI Worldwide, said in an interview last month.
Milwaukee-based Fiserv is another one of the many payment companies that uses AI to analyze payment data and spot fraudulent transactions, CEO Frank Bisignano told analysts in an earnings call in July.
But the criminals are using AI too. FIDO CEO Andrew Shikiar, who leads a global consortium focused on authentication standards, explained last month how fraudsters are expected to become more effective in perpetrating their scams using AI.
With generative AI, the fraudsters can “very quickly create grammatically perfect native language phishing attacks that bring someone to a pixel-perfect replication of the website, which will just make phishing happen at scale,” he said in a December interview. “That’s why it’s super important to put protections in place that prevent someone from getting phished, even if they click on the phishing email, because the phishing emails will be impossible to discern from the real emails.”
As artificial intelligence is used to proliferate scams and fraud, it can also be used to rein them in, said Taira Hall, head of enterprise payments strategy at Citizens Bank.
The emerging technology can analyze large numbers of transactions in an instant and immediately identify red flags, she said.
AI “could create an opportunity to leverage data to better understand customer behavior and recognize patterns, enabling us to identify fraudulent transactions more quickly,” Hall said.
link
