What AI will never change in payments

The payments landscape, much like the broader banking sector, is shifting rapidly thanks to the growing impact of generative artificial intelligence technology (genAI). But as the cliché goes, the more things change, the more things stay the same.

As genAI rewrites the rules faster than many can keep up, some tenets will be immutable – things like trust, and what users (and providers) can expect from the large financial institutions that underpin the payments system.

The growth of genAI itself shouldn’t be a surprise to anybody. A recent United Nations report suggested the value of the global AI market could grow to an astonishing $US4.8 trillion by 2033, about 25 times its worth from just a decade earlier.

For banks, the opportunity is significant. According to McKinsey, successful implementation of genAI’s use cases could “deliver value equal to an additional $US200 billion to $US340 billion” a year. And those use cases are already moving from theoretical to practical.

In the payments space, genAI is playing an increasingly critical role in security and fraud detection, as well as process simplification, for both users and providers. In practice, payments group Remitly has said its AI use lowered its customer support times by three quarters, while fintech Klarna claims its AI assistant is doing the equivalent work of 700 full-time agents.

But through all this change, the fundamentals of financial services have remained largely the same.

Overindexed

The future is defined by change. But it’s easy to fall into the trap of overindexing these shifts, rather than properly considering what can be relied on in the future.

Trust, for instance, is critical in the payments space. And banks will continue to be at the forefront of payments– an Accenture survey found consumers trust their primary bank twice as much as technology providers for products and advice.

Identity plays a big part in that trust. Authentication and authorisation will always be critical elements of that, but how, where and when those elements are applied may evolve in the future.

Today, current processes are designed around human identification – names, addresses, and personal information. But in a genAI-driven future, humans will play a different role than they do today.

The rise of genAI in almost all aspects of trade is driving a concept called agentic commerce – a future where AI ‘agents’ work on behalf of consumers and businesses to find, recommend and complete transactions for products and services. In this world, the know-your-customer principle, or KYC, becomes more about KYA – or know your agent.

Providing access to payment systems and rails will change because these agents will work in a dynamic way. These technologies may be used in ways that appear random compared to today’s demands. The old processes of submitting forms and receiving access to services in three-to-five business days will no longer be fit for purpose.

No matter the origin of the participant, authorisation – and the role of banks – will remain a critical part of the payments system. Industry participants can remain confident that their use of the system will be secure, no matter what happens.

Magnitude

The role of humans in this process will remain important. In an a-commerce world, people will still create and manage these agents, linking them to the organisations, products and services they wish to engage with.

This will have wide-ranging ramifications for the payments space – not least in the area of demand.

As we have seen play out in other areas of the economy, a reduction in barriers is likely to induce higher usage.

Video creation and distribution is a good example. The historic might of Hollywood has been assailed by platforms like YouTube – in viewership numbers, and in revenue.

In much the same way, genAI in banking could lead to an order of magnitude increase in payments. Here, banks will play a critical role in managing this growth, through their proven resilience and scale in other parts of the financial system. Those qualities, important before, will become even more critical as we move forward. This is another thing that genAI will never change.

That it works

It’s an unspoken truth that many new, groundbreaking technologies start off almost like toys. As we saw with the surge in take-up of ChatGPT, much of the early post-launch activity was simply people playing around.

The tool had a simple interface. It was not a new behaviour, and we could all interact with it – and create magic.

The average user didn't care about the diffusion model or latent spaces. They cared about having an image being generated for them when they asked for it. All they cared about was that it worked.

In many ways, it’s the same with how payments technology develops, and where it will develop in the future. Ultimately, for users and providers, the details of the developments will be of minor importance. They’ll only care that it works.

Could things like digital assets, or stablecoins play a role? Could it be something else we are yet to contemplate? All we know is it will be different to today.

What won’t be different is the leading role banks will continue to play in the payments space. And at a time of great change, users can rely on that.

Anuj Mehra is a Director of Strategic Sales & Platforms at ANZ