US consumers embrace AI agents but fear money risk

SEON has published survey findings showing that many US consumers use AI agents regularly but remain reluctant to let them handle money or sensitive information. The results point to a gap between adoption and trust.

The survey of 2,026 US adults found that 68% use an AI agent at least once a week for a personal task. Yet 50% said they were uncomfortable with an AI agent using their credit card details to complete a purchase they had requested.

Responses varied sharply by age group. Among Boomers, 65% said they were uncomfortable with an agent using card details, compared with 52% of Gen X, 47% of Gen Z and 34% of Millennials.

Gender differences also emerged. Women were less comfortable than men: 56% expressed discomfort, compared with 43% of men.

Data sharing

Consumers were more willing to share lower-sensitivity information than financial data. Location data ranked highest at 38%, followed by purchase history and calendar access, both at 34%.

Only 15% said they would share financial information with an AI agent. Younger adults were somewhat more open to doing so, but the numbers remained low: 23% of Millennials and 22% of Gen Z said they were willing, compared with 4% of Boomers.

Almost a third of respondents, or 32%, said they would share no information at all with an AI agent.

Fraud concerns

The findings also highlighted uncertainty about fraud and misuse. Some 43% said they were not confident they could tell whether an AI agent's request for information or account access was legitimate.

Another 39% said they did not trust agents to recognise suspicious or risky situations on their behalf. Confidence in identifying legitimate requests differed widely by generation: 61% of Boomers said they were hardly or not at all confident, compared with 28% of Millennials.

Asked about their main concerns, 20% pointed to a scammer gaining access to the agent or their accounts, while another 20% cited the risk of the agent sharing personal or financial information without permission.

Responsibility

The survey suggested that consumers have clear views on accountability if something goes wrong. Among respondents who expressed an opinion, 65% said the company behind the AI agent should take responsibility for a fraudulent transaction completed by that agent.

Far fewer said responsibility should fall on themselves, at 14%, while 11% chose the merchant and 10% the bank or card provider.

Asked what would increase trust, half of respondents said human support if something went wrong would help most. The ability to reverse mistakes was selected by 42%, while 39% chose strong identity verification.

Even so, nearly a quarter, or 23%, said nothing would increase their trust.

Businesses face a growing challenge in distinguishing between authorised agent activity and fraud. An agent that accesses an account or completes a transaction may be acting on a customer's instructions, but the same behaviour could also indicate misuse by a criminal.

That means identifying the presence of an agent is only the first step in assessing risk. Businesses also need to consider identity, device and behavioural signals to judge who is behind the activity, whether it fits the customer's permissions and whether the wider pattern appears suspicious.

The survey was conducted online among US adults aged 18 and older, and the data was weighted to the US population across nine demographic variables. Questions on guardrails were asked of a random half of respondents, while the question on responsibility for fraud excluded those who answered "not sure".

"Consumers are comfortable letting AI agents help with everyday tasks, but trust drops quickly when money and identity are involved. Half say human support would increase their trust, and they have already decided who is accountable when agent-driven transactions go wrong," said Tamas Kadar, Chief Executive Officer and Co-Founder of SEON.

"Businesses need to tell a helpful agent from a harmful one before money moves. SEON's real-time risk signals show who is behind agent activity, so companies can adopt agents with confidence and keep customers protected," Kadar said.