ACI Worldwide has warned that trust remains a critical barrier to AI shopping agent adoption, after a UK survey found six in ten consumers would stop using an AI agent after one mistake.
The payments technology company said research conducted by YouGov among 2,080 UK adults found a significant gap between consumer interest in AI assistance and willingness to hand over purchasing control.
The survey found that 19% of consumers trust AI assistants to follow rules and make everyday purchasing decisions, compared with 55% who trust a human expert or adviser. A further 69% do not trust AI even when it follows rules they set.
AI shopping agents can search, compare, and, with permission, complete purchases on behalf of consumers. The findings come as retailers, payment providers, and technology platforms bring AI tools closer to checkout, payment, and post-purchase support.
Adriana Iordan, head of merchant and payments intelligence at ACI Worldwide, said: “The findings clearly show that consumers are open to AI helping them shop smarter, but only if they remain firmly in control of both the decision-making and their money. They’re telling us very clearly that they won’t hand control of their finances to an autonomous agent without safeguards. This isn’t a capability gap; it’s a trust and confidence gap.
If the industry wants adoption, it must prioritise control over capability: explicit approvals, hard spending limits, protected payment details and clear accountability when things go wrong.”
The research found that consumers are more comfortable with AI as a support tool than as an autonomous decision-maker. Half of respondents said they trust AI to find the best available price, while 43% trust it to follow spending limits. Trust falls sharply when money, personal data, and problem resolution are involved.
Only 18% trust AI to act in their best financial interest, 17% trust it to keep personal and payment data secure, and 15% trust it to handle problems when something goes wrong.
Resistance rises when AI agents act independently. Seventy per cent of respondents said purchases made without asking would affect their willingness to use an AI shopping agent. Linking the tool to a bank account would affect willingness for 61%, while tracking everything they browse online would affect willingness for 54%.
Responsibility remains unresolved. ACI said 54% of respondents believe the technology or AI company that built the agent should be accountable for refunds when things go wrong. Only 9% blame the retailer, and 3% blame banks or card issuers.
The findings place a clear limit around AI-assisted commerce. Convenience alone is unlikely to overcome anxiety about control, payment security, and accountability. Consumers may accept AI that helps them compare prices or filter choices, but the appetite for autonomous purchasing remains much weaker.
Pressure from AI-influenced shopping is already being felt elsewhere in retail operations. Prime Day activity has shown how AI-supported buying behaviour can add pressure to fulfilment, returns, and logistics planning, particularly when consumers buy faster and compare more widely. ACI’s research adds the trust constraint: AI may help shape discovery and comparison, but payment authority remains a more sensitive step.
The commercial opportunity is still substantial. AI agents could reduce search time, compare offers, identify savings, manage subscriptions, and personalise recommendations. For merchants, they could change traffic acquisition, product information standards, conversion funnels, and pricing visibility.
Agentic shopping also challenges long-standing customer experience assumptions. Retailers have traditionally controlled the storefront, the product page, and the checkout journey. AI agents can sit between the customer and the retailer, filtering options before a consumer reaches a site. That increases the value of accurate product data, fulfilment promises, returns policy, reviews, and customer service reputation.
Payments create a sharper test. Once AI tools are connected to accounts, cards, wallets, or spending rules, consumer protection expectations increase. Explicit consent, transaction limits, dispute processes, refund responsibility, and data minimisation will need to be built into the experience rather than added later.
The finding that 59% of consumers would not trust any organisation to manage AI-powered shopping and payments is especially difficult for the sector. Trust is not automatically transferred from banks, retailers, or technology brands into autonomous commerce. It has to be earned through controls that consumers understand and through remedies that work when something goes wrong.
Early adoption is therefore likely to be cautious. AI agents may first gain traction as advisory tools that shortlist products, monitor prices, or compare specifications, with final approval retained by the consumer. Moving from advice to autonomous purchase will require stronger governance, payment protection, and customer service design.
The race to deploy AI commerce tools is accelerating, but the adoption ceiling will be set by trust, control, and accountability. Companies that treat those safeguards as core product features will be better positioned as agent-led shopping moves closer to checkout.
You must be logged in to post a comment.