Last Updated on November 9, 2026 by Fintech On
Global fraud prevention and anti-money laundering (AML) compliance agency SEON has found that half of all respondents in a recent survey are uneasy about AI agents using their personal card details for payments.
According to an October 2026 SEON survey of 2,026 adults in the US, most adults reported discomfort with AI agents handling card details. However, the results present a clear generational divide in discomfort levels. SEON says 34% of millennials and 47% of Gen Z are uncomfortable, with the numbers rising to 52% among Gen X and 65% among Baby Boomers.
Consumers Worry About Fraud and Mistakes
SEON CEO and Co-founder Tamas Kadar noted that people do not primarily distrust AI agents. 68% of consumers are willing to let these agents handle several general tasks, whether personal or business-related. However, users become uncomfortable when these tasks involve identity or money. Consumers want businesses that use or interact with these agents to quickly distinguish good ones from bad ones before transferring any funds.
Overall, the survey found that the type of data directly influences a respondent’s readiness to share data with AI agents. Notably, only 15% were willing to share financial information. However, 34% said they would share calendar access, another 34% were willing to share purchase history, and 38% were open to sharing location information.
In the event that an AI agent processes a fraudulent or illicit transaction, 88% of the respondents offered varying responses. 10% said the blame is on the card provider or the bank, while 11% blamed the merchant. Another 14% blamed themselves, and 65% said the company responsible for the AI agent should be held accountable.
In the survey, the two most common AI agent fraud issues each concerned 20% of respondents. These issues include unauthorized release of financial or personal information and unauthorized access to an account by a fraudster.
AI Use to Increase Despite Lack of Trust
The survey also revealed that 43% of respondents are not entirely sure they can properly evaluate the legitimacy of an agent’s request. For 39% of the respondents, agents are not trustworthy enough to recognize suspicious transactions. To increase trust, 50% want human intervention when there’s a problem, while 42% want the mistake reversed. For 39%, the solution to the low trust is strong identity verification. However, 23% of respondents do not believe anything can help boost trust.
Despite the lack of trust, more institutions are using AI agents. At the Global Financial Leaders’ Investment Summit in Hong Kong, a State Street bank executive said AI use in finance will increase over the next few years. According to Manoj Bohra, the company’s Chief Data and AI Officer, the bank uses AI in more than 350 scenarios, rising significantly from 150 last year. However, he noted that banks that use AI must not compromise responsibility and must use a regulated operational framework.
