For more than a decade, banks have been busy “going digital”. They built apps, digitised forms, and automated obvious tasks. It worked.

Mobile banking became expected, more processes went paperless, and many core systems were moved to the cloud (or at least parts of them).

Yet even with all that, many banks still haven’t completed that journey. Globally, over 80% of banks have not fully migrated both their core banking and customer-facing digital channels to the cloud.

A recent study also showed that while 91% of banks and insurers have initiated cloud transformation, more than half have moved only minimal portions of their core applications.

In Southeast Asia, 60% of banks report mature or building cloud capabilities, while around 30% are still exploring or planning.

Even so, there’s a growing recognition that doing digital (apps, automation, cloud shifts) isn’t enough.

But Huawei’s Jason Cao steps in with a different frame. He says the banking world is entering its AI era, one where an AI agent might be as much a “customer” of the bank as the human who owns it.

“In the digital time, it’s always you doing the decision but the process itself is automated,” he said.

The AI era, however, will be different. According to Jason, he argues that the decision-maker may well be the agent itself.

Banks, in other words, will need to learn how to satisfy not only customers but also the AI agents making choices for them.

That framing changes the conversation. It’s not just about faster service or cheaper operations anymore. It’s about rethinking who (or what) a bank is actually serving.

From Serving the Top Tier to the Long Tail

Banks have always concentrated their best services on their most valuable customers. Jason is convinced AI will make that model unsustainable.

Banking has long been built on a “20/80” dynamic, where a minority of clients enjoy the majority of services. Jason argues that AI could disrupt this pattern and open the door to more inclusive customer engagement.

“Even if you are a long-tail customer, you can have a dedicated wealth manager. If a bank can make that work, then all the other banks will follow.”

This, he argues, is the real structural change. Hyper-personalisation is becoming a serious competitive edge.

Once a few banks start offering long-tail customers the kind of attention once reserved for high-net-worth individuals, the rest will have to respond or risk losing relevance.

The Urgency to Move

Many institutions are dabbling at the edges of AI, but Jason insists the clock is already ticking. In some markets, especially China, AI agents are already part of day-to-day banking.

The CEO of Digital Finance BU for Huawei believes that once the pioneers can do that, all the other banks will soon follow.

The reason is simple. Costs are falling and capabilities are rising. What used to be prohibitively expensive is now becoming practical.

Jason says the era of “doing digital” was about efficiency. The era of “doing AI”, on the other hand, is about intelligence and reasoning.

Banks that delay the transition may find themselves playing catch-up not with peers, but with entirely new customer expectations set by agent-led interactions.

People and Machines, Together

Jason is careful not to frame this as a story of replacement. The more accurate picture, in his view, is collaboration.

“One person, one team, one position, and multiple agents as a system. This is the model,” he explained. “People, based on their experience, train the agent, and learn how to work with and govern the agent,” he continued.

The point is not to automate humans out of the process but to extend what they can do.

Staff in branches and contact centres will spend less time on repetitive work and more time directing, monitoring and shaping how AI agents perform.

That, in turn, will require new skills and new ways of working.

Jason, however, admits that such a transition won’t happen in a single snap of a finger.

“We have to look forward to what’s coming. It may take years, but we should move to that,” he said.

Making AI Adoption Less Intimidating

Huawei’s new FinAgent Booster, or FAB, is the company’s attempt to lower the barriers. He breaks it down into three promises:

“Easy to use, ready to use, and smooth to use.”

What does that mean in practice?

Fifty scenario-based workflows drawn from banking best practice. More than 150 modular components from partners that banks can plug into their systems.

And the performance to back it up. As of today, it has a 90% accuracy for intent recognition and millisecond-level response times when dealing with customers.

It’s a very Huawei approach, I would say. Modular, engineered, and performance-heavy.

The idea is to make AI adoption feel less like a giant leap and more like building blocks that banks can assemble at their own pace.

The Resilience Story

If there is one word Jason leans on heavily, it is resilience. He repeats it quite often, and he clearly sees it as Huawei’s differentiator.

“Bankers are very smart guys,” he said. “If your products are not good, they will not work with you.”

Despite a difficult six-year period, Huawei’s financial services arm has continued to expand, something the CEO of Digital Finance BU for Huawei links to the trust banks place in its resilience and future growth prospects.

That resilience has been stress-tested. The company has faced global challenges over the past few years, yet its financial services business has expanded.

Today, Huawei works with more than 5,600 financial institutions in over 80 countries, including 53 of the world’s top 100 banks.

Much of the credibility comes from its home market. Every major bank in China, Jason notes, has already undergone large-scale cloud transformation.

That gives Huawei a reference point that international banks pay attention to, especially in Europe, Africa and Latin America, where many are curious about how far they can push their own transformations.

Workforce Shifts Ahead

The deployment of AI agents inevitably raises questions about jobs. Some banks openly talk about reducing thousands of roles.

Jason doesn’t dismiss the possibility, but he emphasises that the bigger story is about reskilling and collaboration.

AI will take over certain functions, but the human role shifts towards oversight, governance and strategic decision-making.

For the industry, that means preparing staff not only for what AI can do today but for how it will evolve over the coming years.

A Sector in Transition

Jason’s tone is pragmatic rather than promotional, stressing that hesitation could cost banks dearly.

The digital chapter was about smoothing processes. The AI chapter is about rethinking who makes the decisions and how those decisions scale across millions of customers.

“In the AI time, the decision-maker will probably be your agent,” he reminds us.

That thought may feel like science fiction, but it is already moulding financial services in China.

For banks elsewhere, the choice is whether to treat it as a warning or as a head start.

Featured image: Edited by Fintech News Singapore based on images by MD.Laik alom mollik via Freepik and Huawei.