Draft poster · the finished cut will appear here
What happened
Reuters reports that a group of major banks warned AI shopping agents are moving faster than consumer-protection standards. The named group includes NatWest, Bank of America, ING, Capital One, Commonwealth Bank of Australia, and New Zealand’s ASB Bank. The warning concerns the path from an agent’s recommendation to an actual payment.
How it works
According to the report, the banks are concerned about agents handling payment credentials, steering shoppers toward payment methods with weaker protections, and making it unclear who is responsible when a purchase goes wrong. They also want clearer disclosure when an AI agent is involved in a transaction.
The banking group proposed stronger data safeguards, interoperability standards, auditable authorization, and clearer accountability. Those proposals describe the missing control surfaces around delegated checkout; they are not a new law or a settled technical standard.
The concern is therefore about recourse as much as fraud. A buyer needs to know which actor chose the item, which credential was used, and which party can correct the transaction when the agent’s decision was wrong.
Why it matters
Delegated checkout needs identity, permission, audit trails, and recourse before it needs another payment button. If the agent chooses the product, route, and credential, the user needs a durable way to see what was authorized and challenge what happened.
What we don’t know yet
Reuters reports the warning and proposals as positions from a banking group, not enacted regulation. The story does not show that every named bank made every example claim, that the proposals have been adopted, or that the banks agree on one technical standard for delegated checkout.
Source: Reuters, Elizabeth Howcroft, September 22, 2026 - original report