There's a gold rush mentality around new artificial intelligence, as
But only 10% of U.S. consumers ranked AI-generated advice as a financial services feature that would be "most interesting" to them, according to research released this week by bank and payment technology company FIS, which surveyed 2,000 consumers and 400 business executives.
"Consumers are not ready," said Melissa Cullen, who leads global strategy, product and commercialization at FIS. "What consumers really want is a better experience. And if AI can deliver that, then the two trends will become aligned."
Companies are much more bullish than consumers are.
More than half of U.S. financial services executives plan to increase investment in machine learning and AI over the next 12 months, according to FIS. Forty-five percent plan to invest in generative AI. Forty-five percent also plan to invest in the metaverse, or web-based virtual worlds that are still in the early stages of adoption in financial services. By comparison, a similar number, or 52% of executives, plan to invest in digital wallets, a relatively mainstream payments and financial services product.
Looking inward
Payment fintech Marqeta this week released two generative AI tools, one for external use and the other for internal. The external tool enables clients or developers to use generative AI to build new payment products on Marqeta's platform.
In the past, Marqeta's platform has been used to enable buy now/pay later and embedded payments, which refers to using payment credentials to access other financial and non-financial services from a single account.
"It's not only consumers where knowledge or interest in AI lags," said Wendy Li, Marqeta's senior vice president of emerging technologies, who heads the payment company's generative AI development. "Even some companies have doubts."
By using Marqeta Docs AI, developers can access the data and other intelligence Marqeta has accumulated by processing and analyzing transactions.
Marqeta Docs AI uses Open AI's Large Language Models to navigate through Marqeta's development site. OpenAI is the Microsoft-backed firm that developed
The internal product involves a generative AI-driven code generator, which is designed to speed internal programming, and subsequently business functions. That product was released a few months ago, ahead of Docs AI.
"It takes time to build that trust. That's why we're using gen AI in its first phase to find the pain points internally," Li said. "Generative AI helps people do their jobs. It can boost productivity."
Payment processors, payment-oriented fintechs and banks are mostly looking at internal use cases given the lower consumer demand, according to Christopher Miller, lead of the emerging technology practice at Javelin Strategy & Research. The primary external use of generative AI is in customer service, as opposed to driving new payment flows, lending decisions, or delivering new consumer-facing products. "Companies are not looking at generative AI as a 'device tool' right now," Miller said.
'A vague sense of dread'
Ninety-four percent of global financial services respondents surveyed plan to increase AI investment in the next year, but the most expected impact of generative AI is internal and business operations improvement, according to research from Lucidworks, a workplace applications company.
"Consumers have high, but vague expectations. Meanwhile, companies have high expectations, but a vague sense of dread," said Mike Sinoway, CEO of Lucidworks.
Because of the mix of fear of falling behind in innovation and fear of a security breach or a "
FIS is developing generative AI for internal use, while collecting data that will inform how it will develop the technology for client and consumer purposes. Most current development focuses on operational functions, such as how to route service queries or provide detailed information to guide transactions.
"It's providing the 'next best action,'" Cullen said, adding the company's clients are interested in solving specific problems such as anti-fraud vetting or managing real-time payment that may lend itself to generative AI usage. "The more specific you can make the use, the better the chance for success in adoption."
AI will also continue to be popular among companies even in the event of an economic downturn, partly because of its potential to reduce overhead. For example, 93% of mortgage professionals say that AI will lead them to reduce headcount over the next three years, according to research from
FIS' research found 84% of consumers and 92% of businesses say high inflation and cost of living are impacting their financial decisions. Ninety-one percent say high interest rates are impacting their ability to innovate and invest.
A downturn in IT investment would separate uses for AI, according to Miller, who said use of AI for payment fraud detection and compliance in a faster processing environment will continue to draw funding over more futuristic or experimental uses. "If companies have to prioritize they'll look at uses that solve problems. Something like advice generation would be less likely," Miller said.
The need to manage expenses while improving user experience for a wider range of users will keep AI as a priority for IT investment, Cullen said.
"The cost of living is high, interest rates are high and there are questions about the stability of the financial services industry. And there's a need to simultaneously serve more generations than at any point in history. Banks don't have the luxury of time," Cullen said.
