Surveys show lack of consumer trust in banks' AI

Consumers don't trust the artificial intelligence banks are offering, according to new research from J.D. Power.

The firm's financial services team examined the topic of artificial intelligence for the first time this month. According to J.D. Power's February banking and payments intelligence report, which surveyed 4,000 retail bank customers, people are reluctant to let AI handle a number of banking tasks, even when a bot could theoretically help them save money by finding good rates or recommending investments. 

Jennifer White, senior director of banking and payments intelligence at J.D. Power, has noticed a misunderstanding among customers about what is being powered by AI. 

"There is a distinction between generative AI and using behavioral data that banks might have on hand to customize experiences," she said in an interview. "Generative AI is a new tool and has a little more distrust along with it than perhaps using big data or algorithmic AI to personalize the experience in a mobile app or get the customer information about how they are spending their money."

How generative AI could reshape financial services in 2024

Bank customers also seem ambivalent about the role AI plays in their lives. Nearly a third believe AI will have limited to no impact on them over the next three years, while 17% believe it will make life worse. Twenty-eight percent think AI will make life better. 

Bank customers have the biggest concerns around using facial recognition to withdraw money from an ATM, with just over one quarter of those surveyed stating that they would never do this; 21% are completely resistant to the idea of tools that automatically change or update investment portfolios. 

There is less resistance to tools that analyze spending, with only 15% of respondents deeming this out of the question. Fourteen percent are resistant to tools that make investment recommendations. A mere 10% would never use an AI-driven tool that helps them avoid fraud.

"When the tool itself has an immediate impact on their security or helps them save time or money, they are more willing to accept recommendations," said White. "If the tool moves money on their behalf like a robo advisor or would require them to use something to get to their money, like facial recognition with an ATM, the alarm bells start to go off." She points out that there is much less resistance to AI-driven tools that make investment recommendations, compared with tools that automatically change or update a portfolio. 

"They stop short of letting AI actually move their money," she said.

Still, a third of customers would use tools that change investment portfolios when the technology is proven, and 29% feel the same about facial recognition software at ATMs. 

Even more would try these technologies immediately or already use them. That means banks have a discrepancy to solve.

"Banks must repeatedly communicate the benefits of these tools so customers are more willing to trust them and more willing to use them," said White. Moreover, nearly two thirds of those surveyed answered "somewhat" to the question of "how much do you believe using AI in financial services puts you at greater risk for fraud and security breaches?" Twenty percent answered "extremely." 

These findings are consistent with results from two other recent reports.

On Tuesday, Commonwealth, a nonprofit that focuses on financial security for low-income Americans, released a report on emerging technology and AI. It found that concerns about security and privacy are significant barriers to consumers using emerging technologies in finance. For instance, 63% of the 3,000 respondents — who encompass both households living on low- to moderate-incomes and those with higher incomes — cited security as a top concern when using a financial chatbot. Fifty-three percent named privacy. Comparatively, 49% mentioned unreliable responses. 

People are also reluctant to turn to an AI-driven bot for some banking activities, according to a report published on Thursday by marketing technology company Vericast. It found that 71% of consumers it surveyed prefer to speak with a human agent versus bot for financial advice, while 64% feel the same about fraudulent transactions.