AI Shopping Agents Trust: Why Banks Just Sounded the Alarm

Six of the world's largest banks just told regulators something Silicon Valley did not want to hear: AI shopping agents are moving faster than the guardrails meant to protect consumers. Using AI agents for online shopping could increase the risk of scams, fraud and data-privacy breaches, banks including NatWest and Bank of America said on Tuesday, as they set out principles for developing the technology. The coalition, which also includes ING, Capital One, New Zealand's ASB Bank, and Commonwealth Bank of Australia , is not trying to slow the technology down. It is trying to make sure someone builds the trust layer before consumers get burned.

This matters because the shift is already happening at a pace few executives have fully internalized. British retailer John Lewis said in September that searches originating from AI agents had risen to 2.5% from 0.3% a year earlier, with the trend accelerating. That is an eightfold increase in twelve months, and John Lewis called the growth "exponential, and it's all age groups." For Fortune 500 leaders, this is the equivalent of watching mobile commerce accelerate in 2010 while dismissing it as a niche behavior.

Matt Britton, founder of Suzy and author of Generation AI, argues that most brands are reading this news the wrong way. The instinct is to treat the bank warning as a reason to wait for regulatory clarity before investing in agentic commerce infrastructure. Britton's counterpoint is direct: waiting is the riskiest strategy on the table, because AI agents that cannot parse a brand's data will simply route customers to a competitor that made itself legible to the machine.

The banks' report is not an indictment of agentic commerce. It is a signal that AI shopping agents trust has become the defining battleground of the next commerce cycle, and the brands that treat transparency as a product feature rather than a compliance checkbox will own the customer relationships that used to belong to search engines and marketplaces. This is the framework Matt Britton has been building keynotes around for Fortune 500 boards preparing for a checkout experience their brand no longer controls.

What the Bank Warning Actually Says About Agentic Commerce Risk

The report from NatWest, Bank of America, and their peers is more specific than a generic caution about new technology. The banks warned that agentic shopping has "risks spanning transparency, safety, privacy & data, choice, and interoperability" that increase "as greater autonomy is given to AI agents." That framing matters because it identifies the exact failure points brands and platforms need to solve, rather than issuing a vague call for caution.

Two risks stand out as immediately actionable for any business building a commerce stack. First, the report highlighted risks including AI agents requesting customers' card details and entering them directly into websites, or steering users towards payment methods that offer weaker protections. Second, the underlying fear is one of accountability. Banks are concerned that AI agents may buy the wrong thing or spend too much, or even worse, lose their money to scams and fraud, and customers are not sure whether they will be protected or who they will need to go to if things go wrong.

The banks are not asking regulators to shut this down. They plan to discuss a series of proposals with policymakers, including requiring disclosure when an AI agent is involved in a transaction, greater transparency over how AI agents make decisions, and safeguards to protect customer data. That is a blueprint, not a moratorium, and Matt Britton tells corporate audiences that the companies who build ahead of those disclosure standards will be the ones setting the terms rather than scrambling to comply with them.

Agentic Commerce Brand Strategy: Why "Agent Legibility" Is the New SEO

For two decades, brands optimized for how humans search. Now they need to optimize for how machines decide. Matt Britton calls this shift "agent legibility," meaning the degree to which a brand's product data, pricing, reviews, and policies can be cleanly read and trusted by an AI system making a purchase decision on a consumer's behalf.

John Lewis offers the clearest live case study in the market today. The retailer's AI shopping push is being backed with a significant investment as part of its broader multi-year transformation programme, and it has extended its partnership with agentic commerce platform Commercetools. The goal is explicit: the retailer's products will be served to customers looking for inspiration on AI platforms such as Google Gemini and ChatGPT.

That kind of investment is not optional anymore for brands with real digital revenue. Around 60% of John Lewis orders now take place online, meaning even a relatively small change in how consumers discover products can become commercially significant if the trend continues. Britton's argument is that this dynamic scales across every category, from real estate to financial services, which is why his keynote content increasingly draws direct lines between property discovery and financial services trust as parallel agentic commerce battlegrounds.

Building agent legibility into a commerce stack requires four concrete moves:

• Structuring product data in formats AI agents can parse without ambiguity, including specs, availability, and return policies.
• Publishing clear, machine-readable disclosure of pricing and fees so agents do not default to guesswork that erodes trust.
• Auditing checkout flows for the exact card-handling and payment-routing risks the banks flagged in their report.
• Establishing a visible accountability chain so customers and agents alike know who resolves a dispute when something breaks.

Brands that skip this work are not being cautious. They are becoming invisible to the exact discovery layer their customers are increasingly using.

AI Agent Fraud Consumer Trust: The Data Behind the Alarm

Skeptics might argue the bank warning is overblown because adoption is still small. The numbers suggest otherwise, and the trajectory is what should concern executives more than the current share. Research from PYMNTS Intelligence showed that half of American consumers reported completing a retail purchase with some form of AI help, yet only about one in four said they would be comfortable handing both the shopping and the payment entirely over to an AI agent.

That gap between usage and trust is the entire opportunity. A separate global study found trust in AI is closing in on traditional channels: 36% of consumers now trust AI to influence their purchases, nearly matching the 38% who rely on in-store associates. Yet consumer anxiety remains concentrated in specific, solvable areas. Payment security is the main worry for nearly one in three shoppers (32%), followed by privacy (26%), potential mistakes (18%), and loss of control (17%).

The fraud mechanics behind that anxiety are not hypothetical. Security researchers have noted that OpenAI's Chief Information Security Officer stated in October 2025 that prompt injection remains an unsolved security problem , meaning bad actors can manipulate agents into taking actions the consumer never authorized. On the identity side, synthetic identity fraud now accounts for 85% to 95% of all fraud losses in the identity fraud category in the US , and generative AI has compressed the time it takes criminals to manufacture those fake identities.

Matt Britton frequently tells finance and retail audiences that this is precisely why speed matters more than perfection. Brands do not need to solve every fraud vector before entering agentic commerce. They need to demonstrate, visibly and consistently, that they are building toward the disclosure and accountability standards the banks are now pushing regulators to adopt.

Banks AI Shopping Regulation: What Comes Next for Policymakers and Brands

The banks made clear this is the opening move in a longer policy conversation, not a finished framework. The report arrives as U.S. regulators are also intensifying their attention on AI in financial services, and banking examiners have incorporated AI oversight into routine bank examinations, pressing lenders on how they govern the technology across higher-risk functions, though no AI-specific rules exist yet. That regulatory vacuum is exactly the window Matt Britton tells clients to move into now, rather than treating it as a reason for delay.

History supports his read. Every major commerce shift, from e-commerce in the late 1990s to mobile checkout in the 2010s, saw regulation trail adoption by years. The companies that built trust infrastructure ahead of the rules, rather than waiting for mandates, captured disproportionate market share once the rules eventually arrived. Britton's keynote work through his AI-focused speaking platform centers on this exact pattern: first-mover trust compounds, and it rarely gets redistributed once regulators catch up.

Retailers are not waiting idly either. Retailers are actively trying to optimize their product data to sway how chatbots generate recommendations , which means the competition for agent visibility has already started even without a regulatory framework in place. Brands that treat the bank warning as a green light to build responsibly, rather than a red light to freeze, will define what "trustworthy agentic commerce" looks like before any regulator writes it into law.

Key Takeaways for Business Leaders

• Audit your checkout flow now for the exact card-handling and payment-routing risks banks have flagged, before regulators mandate fixes.
• Invest in agent legibility by structuring product data, pricing, and policies so AI systems can parse them accurately on your customers' behalf.
• Publish clear disclosure standards around AI involvement in transactions ahead of policy mandates to build first-mover consumer trust.
• Track agent-originated traffic the way John Lewis does, since even small percentage shifts signal where discovery is heading next.
• Treat trust as a growth strategy, not a legal obligation, because the gap between AI usage and AI trust is where market share will be won.

Frequently Asked Questions About AI Shopping Agents Trust

Why are banks warning about AI shopping agents now?

Banks including NatWest, Bank of America, and Capital One warned that agentic commerce technology is advancing faster than consumer protections, creating risks around fraud, data privacy, and unclear accountability when transactions go wrong. They are pushing regulators to establish disclosure and transparency standards before adoption scales further.

What is agentic commerce risk for consumers?

Agentic commerce risk refers to the exposure consumers face when AI agents make purchasing decisions or handle payment details on their behalf, including scams, incorrect purchases, overspending, and confusion about who is liable if fraud occurs. Banks specifically flagged AI agents entering card details directly and routing payments through less secure methods.

How should brands respond to declining trust in AI shopping agents?

Brands should build transparency directly into their commerce infrastructure, including clear disclosure of AI involvement, machine-readable product data, and visible accountability for disputes. Matt Britton advises companies to treat this as a competitive opportunity rather than waiting for regulation, since the brands that build trust first will capture the customer relationships agents control.

Will AI shopping agents replace traditional e-commerce?

AI shopping agents are unlikely to fully replace traditional e-commerce in the near term, but their share of product discovery is growing quickly, as shown by John Lewis reporting an eightfold increase in AI-originated searches within a year. Brands that fail to optimize for agent discovery risk becoming invisible in an increasingly AI-mediated shopping journey.

The Trust Arms Race Has Already Started

The bank warning is not a reason to slow down. It is confirmation that AI shopping agents trust has become the metric that will separate the retail winners of the next decade from the brands still waiting for permission. Matt Britton has spent years telling Fortune 500 audiences that consumer behavior moves faster than corporate governance, and this moment proves the pattern again.

Brands now face a binary choice: build the transparency and legibility that agents and regulators are demanding, or get disintermediated by competitors who move first. Matt Britton unpacks this shift, along with the data behind it, in keynotes built specifically for leadership teams navigating agentic commerce, consumer trust, and the AI-driven future of retail.

Explore how Matt Britton can bring this research to your next leadership offsite or industry conference through his speaker platform, or dig deeper into the underlying trends shaping this shift in his book Generation AI. For ongoing analysis of consumer behavior and AI adoption, subscribe to The Speed of Culture podcast, and see how Suzy's consumer intelligence platform tracks these shifts in real time.