UK consumers trust AI less with payments than health

UK consumers are less willing to trust artificial intelligence with payments than with health or legal questions, according to research by Enfuce. Among 10 everyday tasks tested, payments ranked last for trust in AI.

The survey of 2,000 UK consumers found that 37% would trust AI to help with a health symptom and 38% with a legal question, compared with 22% who would trust it to make or manage a payment.

Use of AI across these areas also differed sharply. Among people who already use AI, 53% said they had used it for general search and 26% for health symptoms, while 10% had used it for savings and investments and 8% for payments.

The findings point to a specific trust problem in payments rather than a broader rejection of AI. General search ranked highest for AI trust at 63%, followed by navigation at 57% and recommendations at 56%.

Understanding gap

The research also highlighted weak consumer understanding of how card payments work. Nearly half of respondents, 48%, thought only one or two companies were involved in processing a card payment, while 27% said they understood well how their payments were kept secure.

That lack of visibility appears to shape which organisations people trust. Banks and card networks were trusted by 88% of consumers to keep their money secure, compared with 41% for fintech firms and payment apps.

Familiarity also influenced confidence in payment methods. Cash and direct debit each scored 82% for trust, while contactless payments scored 64% and digital wallets 57%.

When asked who would be mainly responsible if money was lost to a scam, no single party was identified by more than a quarter of consumers. Some 23% pointed to the platform where the scam originated, 21% blamed the consumer, 15% named the consumer's bank, and 15% said they did not know.

That uncertainty was reflected in what consumers said would improve trust. Some 30% said knowing they would get their money back if something went wrong would make the biggest difference, ahead of strong fraud protection at 19%, personal data protection at 8%, and provider reputation at 6%.

Fraud concerns

The survey found resistance to AI even when it is used as a safeguard rather than a payment tool. While 95% of respondents said they would accept at least one extra security measure in exchange for more secure payments, only 19% said they would accept AI-led fraud detection as that safeguard.

Consumers were more comfortable with measures they could identify directly. Some 47% said they would accept requiring approval before a payment, 43% accepted additional identity checks, and 36% cited reimbursement protection and access to human support.

Asked how they would feel if fraud detection was handled by AI rather than a human team, 45% said they would feel uncomfortable, and 23% said they would feel comfortable.

Nine in 10 respondents said they had at least one concern about AI being used to detect payment fraud. The main worries were AI making mistakes, cited by 51%, followed by data privacy at 39% and cybersecurity risk at 35%.

The pattern suggests consumer hesitation is tied not only to AI itself but also to the structure of the payments system, where multiple intermediaries sit behind a transaction and responsibility is not always clear to the person making the payment.

Denise Johansson, Chief Executive Officer and Co-founder of Enfuce, said the research showed the industry could not assume trust would follow as AI takes a larger role in the background of payments.

"We spent decades making payments invisible: tap, pay, move on. Yet our research shows people are more willing to trust AI with a legal question than with a payment. Our industry needs to take that hesitation seriously. Customers should be able to trust that their money is protected, that someone is accountable for it, and that there is somewhere to turn when something goes wrong. Payments can stay invisible. Trust cannot."