SEON has found that 50% of surveyed US adults are uncomfortable with AI agents using their card details, even as 68% use agents weekly.
Fraud prevention and anti-money laundering compliance provider SEON surveyed 2,026 US adults online in October 2026, with results weighted to the US population and a credibility interval of ±3 percentage points. While 68% of respondents use an AI agent at least once a week for a personal task, comfort drops when money is involved. Discomfort with an agent using card details to complete a requested purchase varies by generation, from 65% of Baby Boomers and 52% of Gen X to 47% of Gen Z and 34% of Millennials, and is higher among women (56%) than men (43%).
Limited willingness to share financial data
Respondents’ willingness to share data with AI agents depends on the type of information. Location (38%), purchase history (34%), and calendar access (34%) are the most widely accepted, while only 15% would share financial information. Younger adults are more willing, though fewer than one in four would share financial details, at 23% of Millennials and 22% of Gen Z compared with 4% of Boomers, and 32% of all respondents would not share any information with an agent.
Asked who should be responsible if an agent completes a fraudulent transaction, 65% of the 885 respondents who gave an opinion named the company behind the AI agent, compared with 14% who said themselves, 11% the merchant, and 10% their bank or card provider. The most common fraud concerns were a scammer gaining access to the agent or the user’s accounts (20%) and the agent sharing personal or financial information without permission (20%).
A confidence gap in spotting fraud
In the survey, 43% of respondents said they lack confidence in judging whether an agent’s request is legitimate, and 39% do not trust agents to recognise suspicious situations on their behalf. Among Boomers, 61% said they are hardly or not at all confident they could tell whether a request was legitimate, compared with 28% of Millennials. Human support when something goes wrong was the main factor that would increase trust (50%), followed by the ability to reverse mistakes (42%) and strong identity verification (39%), while 23% said nothing would increase their trust.
Tamas Kadar, CEO and Co-Founder of SEON, stated that consumers are comfortable letting AI agents help with everyday tasks, but trust drops quickly when money and identity are involved, and businesses need to distinguish a helpful agent from a harmful one before money moves. According to SEON, detecting that an agent is involved is only the first step, and businesses need to assess agent activity alongside identity, device, and behavioural signals to determine who is behind it and whether it matches the customer’s permissions.
