Educational research only — not investment advice.
AI shopping agents could change online commerce much faster than many consumers expect.
Instead of simply recommending a product, an AI agent could:
- search across stores
- compare prices
- choose an item
- fill in checkout details
- make a purchase
This new model is often called agentic commerce.
But banks are warning that it also creates a new question:
Who is responsible when the AI buys the wrong thing—or sends money to the wrong place?
What Is an AI Shopping Agent?
A normal chatbot gives advice.
An AI shopping agent can take action.
For example, a user might say:
“Find me the best laptop under $1,000 and buy it.”
The agent could search several stores, compare products and complete the transaction within rules set by the user.
OpenAI, Google, Anthropic and Meta are all developing AI systems that can increasingly perform tasks rather than only answer questions.
That could make online shopping much faster.
It also creates new financial risks.
Why Are Banks Worried?
Major banks including Bank of America, NatWest, ING, Capital One, Commonwealth Bank of Australia and ASB Bank have warned that AI shopping is developing faster than current consumer protections.
One concern is payment information.
An AI agent may need access to:
card details + addresses + purchase history + personal preferences
If that information is mishandled or stolen, fraud becomes much easier.
Banks also worry agents could direct users toward payment methods with weaker protections.
Fraud Could Become Harder to Understand
Imagine an AI agent buys a fake product from a scam website.
Who is responsible?
The consumer?
The AI provider?
The bank?
The merchant?
That question is still not fully settled.
Banks say consumers are uncertain about who protects them when something goes wrong.
This becomes especially important when AI systems make thousands of decisions automatically.
The Opportunity Is Still Huge
Despite the risks, consumers are already beginning to use AI for shopping.
British retailer John Lewis said searches coming from AI agents increased from 0.3% to 2.5% of traffic in one year.
Payment companies also see major potential.
Visa found that only 23% of U.S. consumers currently trust generative AI to handle payments, showing that trust—not technology—may be the biggest barrier to adoption.
Mastercard is already building tools designed to let merchants support AI-powered product discovery and authorized purchases.
What Needs to Change?
Banks are pushing for stronger rules before agentic commerce becomes mainstream.
Possible safeguards include:
clear AI disclosure
Consumers should know when an AI agent is involved.
strong payment authorization
Agents should only spend within limits approved by users.
better data protection
Card details should not simply be passed between unknown systems.
clear responsibility
Consumers need to know who handles refunds and fraud disputes.
These protections could determine how quickly people become comfortable allowing AI to spend their money.
Why This Matters for Markets
Agentic commerce could eventually affect several industries:
- banks
- Visa and Mastercard
- online retailers
- payment processors
- advertising platforms
- AI companies
The opportunity is enormous because AI agents could become a new layer between consumers and merchants.
But whoever controls that layer may also control:
product discovery + payments + customer data
That makes trust extremely valuable.
What Should Investors Watch?
Watch AI shopping adoption, payment fraud, bank regulation, Visa and Mastercard initiatives, and retailer traffic from AI agents.
The key question is simple:
Will consumers trust AI enough to let it spend money for them?
If the answer eventually becomes yes, agentic commerce could become one of the biggest changes to online shopping since the smartphone.
But security and payment protection will need to improve alongside it.
Track Technology Trends With TradingSimuLab
TradingSimuLab’s Trend Detector and Risk tools help users study changing technology themes, market momentum and emerging risks.
For more quantitative market research and educational trading tools, sign up to TradingSimuLab.
TradingSimuLab is for educational and research purposes only and does not provide investment advice.