But those examples of how AI is affecting companies do not necessarily mean banks in Southeast Asia will follow the same path overnight. They do, however, make the trust question harder to avoid.
If AI is powerful enough to change how banks hire and structure their teams, it also needs to be reliable enough to sit inside real financial operations.
Like many other companies, banks and fintechs also want the upside of artificial intelligence. AI has proven that it could help with faster onboarding, better fraud detection, lower operating costs and more personalised customer experiences, all of those shiny things.
But what they usually missed out on is at the flipside of the coin, where the question of knowing whether the systems behind those gains can be monitored and defended when something goes wrong.
Leo Li, President of International Business at Ant Digital Technologies, sees that hesitation clearly.
“From my observation, the two biggest concerns are ROI and risk,” he said.
Those two words capture why many banks remain cautious.
AI can be useful, but usefulness alone is not enough in a sector where every new system must answer to compliance, audit and customer trust.
Banks Are Starting at the Edges
Leo Li
“Most banks are slowly utilising AI from the outside in,” Leo said.
What Leo meant by this is that he views AI adoption in banking as a gradual movement from the outer parts of the organisation toward the core.
On the one hand, things like customer service, marketing, onboarding and operational support tend to come first because the risks are easier to manage.
Core banking on the other, sit much closer to the institution’s risk centre, where mistakes are harder to explain and more expensive to fix.
Across the market, this creates a strange mix of momentum and caution.
Banks are already testing AI assistants, chatbots, fraud tools and even internal productivity systems, while still keeping a tighter grip on use cases with heavier financial or regulatory consequences.
A pilot can show that something works under controlled conditions but daily use requires more confidence because the bank must be able to explain the system’s role when a decision is questioned.
Leo said banks usually approach AI with a practical question in mind:
“Does it create enough value to justify the risk?” he asks.
The Accountability Problem Has Not Gone Away
That is where Leo draws a clear line between capability and responsibility.
“AI is a tool. It cannot take responsibility. In the end, humans and institutions still carry that responsibility,” he said.
The distinction matters because AI-supported decisions can affect access to a lot things.
A faster model may improve efficiency, and a sharper system may reduce some risks, but the institution still answers to customers, regulators and its own board.
Banks and financial regulators especially within Southeast Asia are also moving closer to the issue that AI poses.
Neighbouring country, Singapore’s Monetary Authority of Singapore has set out FEAT principles covering fairness, ethics, accountability and transparency in the use of AI and data analytics in financial services.
Leo’s framework for financial AI sits in that same territory. He said AI used in finance must be explainable, auditable and traceable.
“When we deploy AI in the financial industry, AI must be explainable, auditable and traceable,” he said.
A bank needs to know why a customer was rejected, why a transaction was flagged, how a model reached a recommendation and whether the process can be reconstructed later.
Without that discipline, AI risks becoming another opaque layer inside an already complex system.
SMEs May Feel the Impact Differently
The same accountability question becomes more interesting when AI moves beyond the bank itself and begins to shape how financial institutions understand smaller businesses.
Leo sees that shift playing out not only inside banks, but also among the SMEs they serve. In Southeast Asia, that is not a small market.
Leo pointed to the rise of what he called OPCs, or one-person companies, as one signal of how AI could change business formation.
In China, he said, some individuals are already using AI agents and tools to manage work that previously required several people.
“One person can utilise agents and AI tools to build the whole system for one company,” he said.
Southeast Asia’s SMEs will not all move in that direction. Many still face familiar constraints around financing, digitisation and resilience.
Still, AI could give smaller teams more capacity to manage operations, reach customers and generate clearer signals of how their businesses are performing.
That matters for banks and fintech lenders because smaller businesses may not look the same as they did before.
If AI changes how these firms operate, financial institutions may need to rethink what a small business looks like in the first place, while also still keeping the accountability standards expected in regulated finance.
Local Data Becomes the Differentiator
“Data belongs to the region. Malaysia’s data, for example, is unique,” Leo said.
The point connects back to SMEs because understanding how smaller businesses operate will depend on more than access to AI models alone.
A lender trying to assess an AI-enabled merchant in Malaysia cannot rely only on assumptions built for another market.
It needs to understand how local businesses receive payments, manage cash flow, verify customers and respond to seasonal demand. Those details are often where risk becomes visible.
Leo said AI capability usually depends on data, along with computing power and algorithms.
The latter two have become easier to obtain as cloud infrastructure improves and large models become more widely available. The harder part is the first one, data, because it reflects how people and businesses actually behave in each market.
Many financial institutions already have that information sitting inside their own systems.
Leo described banks as sitting on a “gold mine”, but the value of that data depends on whether institutions can use it responsibly and with enough governance.
Privacy computing is one area he believes could help.
It allows institutions to draw value from data without unnecessarily exposing personal information, which becomes important when banks are trying to understand SMEs that may not fit neatly into traditional credit models.
Better local data could help financial institutions see how smaller businesses are changing. Poor governance, however, could turn the same opportunity into another trust problem.
Trust Has to Be Earned in Production
Better local data can help banks understand customers and SMEs more clearly, but the real test begins when that data is used inside live financial services. The technology has to work reliably, and the institution has to know what to do when it does not.
For Leo, trust only becomes meaningful in production.
“Trust is based on outcomes and results. It is not built on marketing or slogans,” he said.
Banks do not evaluate technology partners only on product claims or global credentials. They need providers that can respond quickly, adapt to local requirements and support systems once they are exposed to real customers, transactions and risk decisions.
Ant Digital Technologies, the technology arm of Ant Group, sits in that part of the conversation through infrastructure, AI infrastructure, mobile platforms, digital identity, AML and risk management, with credit-related services depending on local licensing requirements.
In Malaysia, the company not only work with clients from the banking and fintech side like RYT Bank, Kenanga Investment Bank Berhad (KIBB) and TNG Digital, they also worked with SP Setia, a property developer within the country.
Its local presence has also grown through ZOLOZ, Ant Digital Technologies’ flagship AI product, which launched a Malaysia operations hub in 2026 to strengthen local service capabilities, response times and on-the-ground processing for Malaysian customers.
Leo said Ant Digital Technologies sees international markets as a long-term commitment, with Malaysia playing a role as a talent and resource centre.
Local partners lead, while the company supports with technology, infrastructure and operating experience.
Trust in financial AI therefore has to be proven close to the market. Banks may work with global technology providers, but customers experience the outcome through the local institutions they already know.
AI Strategy Needs the CEO
Early AI work may sit with technology teams, but they cannot carry the transformation alone.
Leo believes the CEO has to carry the responsibility because AI changes how the organisation makes decisions, manages risk and responds to customers.
Banks may find that integrating the platform is the easier part. Changing the organisation around it will take longer.
“AI transformation should be every CEO’s top priority,” he said.
Featured image: Edited by Fintech News Indonesia based on an image by Frolopiaton Palm via Magnific.
This article first appeared on Fintech News Singapore.
