6 Global banks collaborate on “principles for trusted agentic commerce”

Consortium of banking giants from the US, UK and Australia publish joint principles on responsible development of agentic commerce, safety and security

Add bookmark

CX Network
09/23/2026

Low angle view of skyscrapers against a clear sky in a bustling urban area.

ASB Bank, Bank of America, Capital One, Commonwealth Bank of Australia, ING Group and NatWest Group have jointly published a shared set of principles they say will support the development of "trusted agentic commerce".

The consortium is looking to establish a framework to allow customers and merchants to "retain choice, control and flexibility" in how they pay and get paid in a future where machine-to-human and machine-to-machine commerce becomes the norm.

The banks are focused on five areas: transparency, safety, privacy and data, choice and interoperability. The paper in which their framework is published is "intended to build on a wider conversation, with the participating banks inviting other banks and institutions across the payments ecosystem to engage with the principles, share different perspectives and work together on developing a blueprint for applying the principles in practice".

Don't miss any news, updates or insider tips from CX Network by getting them delivered to your inbox. Sign up to our newsletter and join our community of experts.

Do customers trust agentic commerce?

Yes and no.

The number of consumers using personal AI assistants to research, compare and shortlist products and services continues to climb. Figures published by Adobe show retail sites in the US saw a 393 percent increase in traffic from AI sources in Q1, 2025.

According to CX Network's 2026 research into the state of CX, AI-first customer journeys are front of mind for CX practitioners. When asked to select three CX trends and three customer behaviors influencing their work between 2026 and 2030, AI-first customer journeys emerged as the third most selected response to both questions.

Intent among shoppers who use AI assistants is generally much higher, however, the conversion from discovery to sale is less strong. Figures published earlier this year by NRF and Salesforce concluded that although 45 percent use their AI assistants for product discovery, fewer than 12 percent said they had completed a transaction end-to-end in a third-party conversational interface.

In March, OpenAI withdrew the Instant Checkout feature on ChatGPT, which lets shoppers complete a purchase without leaving the chat interface.

Safe and secure payments are widely recognized as the primary barrier to wider adoption.

As a concept, machine-to-human and machine-to-machine commerce has been talked about in CX for years. In addition to the new payment rails and new legal frameworks required to verify the authority of AI agents, the shift to machine customers also demands new ways to ensure the safety and security of payments.

According to Michael Nevski, who most recently held the role of director of global insights for Visa's Business and Economic Insights team: "Consumers already grasp the appeal, efficiency, personalization, reduced effort and potential savings [of agentic commerce]. What they are waiting for is not better technology. They are waiting for a trustworthy system around the technology."

Nevski will present his research on this topic during the virtual event Mastering CX: AI Agents in October 2026. Registration is free via this link.

How will the framework on agentic commerce payments nurture customer trust?

In its introduction, the paper outlines that trust "will be critical to support adoption at scale" and that all market participants have a role to play in "unlocking the opportunities of agentic commerce, while addressing outstanding gaps, risks and interoperability challenges".

It then explores the five focus areas in detail and sets out specific actions for each area that will support safer agentic commerce. It's important to note the contents of the paper are voluntary and non-binding, but are still surprisingly light on detail, particularly when it comes to which institutions and stakeholders should take responsibility for the many things that still need addressing.

The full paper can be accessed via this link.

Transparency

The recommendations start strong. On this point, the paper states that all parties involved in agentic commerce "should be aware when an AI agent is helping to make or facilitate a transaction, and on whose behalf they are acting".

This should also extend to consumers and merchants having transparency over how AI agents prioritize options and make decisions, including any sponsored options, and how their data is collected and used.

Safety

To support safe agentic commerce, the paper recommends service providers be "discouraged" from "unsafe practices".

However, the strongest recommendation here is that consumers should retain "appropriate control over the authority they grant to AI agents, including the ability to view and manage delegations".

Providers, on the other hand, should "use secure and auditable methods to enter payment credentials and authorize purchase intents and payments" and "work together to prevent, detect, disrupt and respond to safety risks".

With regards to dispute resolution, the consortium suggests this be "fair and efficient, with liability outcomes reflecting where risks and errors are introduced".

Privacy and data

Here is where valid principles begin to run out of steam.

To support data governance in agentic commerce, the consortium says service providers should meet appropriate standards and maintain auditable records. They suggest service providers should also limit data collection/access and that consumer and merchant consent should determine additional data sharing.

In short, it heavily follows the data protection regulations already enacted across jurisdictions such as the EU.

Choice

This one has some broader implications for the market, hinting that monopolies should not be tolerated and that consumers should always have alternatives.

Service providers should have the freedom to choose which organizations they engage with and to be able to make those decisions on merit rather than interoperability or dominance.

It also states "consumers should have the option to share their identity with merchants if they choose".

Interoperability

Aside from the obvious calls, these recommendations extended to considering expanding interoperability and connectivity to "additional areas where common approaches deliver benefits for consumers and merchants". No examples were given, however.

How are organizations building trust around agentic commerce today?

Announcing the new joint framework Jonathan Oram, executive general manager of corporate banking and payments at ASB Bank said: "Agentic commerce has the potential to transform how people and businesses interact with the economy, but its success will ultimately depend on the strength of the underlying payments infrastructure.

"New Zealand is already undertaking significant payments modernization, and we see a clear opportunity to ensure new AI-enabled payment experiences are built on trusted foundations that support customer control, competition, security and interoperability from day one," Oram continued.

In April, American Express and Visa launched new payment tools and frameworks to facilitate secure payments, with Amex also unveiling Amex Agent Purchase Protection™, which gives human customers industry-first protections in instances of AI agent error, but not fraud.

"AI agents are beginning to reshape how people discover products and services, plan travel and dining, and make purchases," said Luke Gebb, EVP and head of global innovation at American Express in a press release. "As these capabilities evolve, card members and merchants will expect the same level of trust and security that they always relied on from American Express. The ACE Developer Kit enables this in AI-powered commerce."

In its press release, the consortium said the next step is to "detail in a subsequent paper how the principles can be implemented". It welcomes engagement with customers, industry and stakeholders as this work progresses.

The principles paper can be accessed in full via this link.