Fintech · July 21, 2026
HSBC Joins EPAA Agentic AI Payments Working Group in APAC
EPAA has launched an AI & Agentic Payments Working Group with HSBC as a founding member, aiming to set safety standards and interoperability frameworks for autonomous AI-initiated transactions across Asia-Pacific.
What happened
The Emerging Payments Association Asia (EPAA) has launched a dedicated AI & Agentic Payments Working Group, with HSBC confirmed as a founding member. The group brings together banks, payment networks, fintechs and technology platforms with the stated aim of establishing the standards, guardrails and interoperability frameworks needed to make agentic commerce — where AI systems initiate and complete transactions autonomously on behalf of users — operate safely and at scale across the Asia-Pacific region.
The working group's formation signals a formal industry shift from experimenting with AI in payments to governing it. Rather than individual institutions building proprietary approaches, the EPAA model convenes competitors around shared standards, a pattern that has historically accelerated adoption in payments infrastructure while reducing systemic risk.
Why it matters
Agentic AI represents one of the most consequential shifts in the customer experience of financial services in a generation. When an AI agent can research, decide and pay — without a human approving each step — the traditional CX touchpoints of choice, confirmation and consent are compressed or removed entirely. For service designers, this is not simply a technology question; it is a fundamental redesign of the customer's role in their own financial journey. The working group's focus on safety and scale suggests the industry is already grappling with the trust deficit that autonomous spending could create if it goes wrong.
From a behavioural economics perspective, agentic payments remove friction — which is usually a CX goal — but friction also serves as a deliberate pause point that protects customers from regret, error and fraud. Designing for a world where AI acts on standing instructions requires new consent architectures, explainability standards and recovery mechanisms. The institutions that get this right will build durable customer trust; those that treat it purely as an efficiency play risk significant backlash when the first high-profile autonomous payment error occurs.
The Renascence take
Most commentary on agentic payments focuses on speed and automation. What the EPAA initiative quietly acknowledges is that the real design challenge is not making AI transact faster — it is making customers comfortable enough to delegate financial decisions to a machine in the first place.
The working group's value will not be measured in the standards it publishes but in whether those standards encode meaningful customer control — the ability to set boundaries, review decisions and recover gracefully when something goes wrong. Behavioural research is clear that people accept automation when they feel they can override it; the moment that sense of agency disappears, trust collapses disproportionately. Customer-obsessed operators in APAC should be at this table not to shape payment rails, but to ensure that the human experience of delegating to AI is designed with the same rigour as the technology itself. If CX voices are absent from working groups like this one, the resulting standards will be engineer-led — and customers will feel it.
Sources
This briefing was written by the Renascence newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.
More in Fintech
Stay ahead of CX
Get the signal, not the noise.
The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.