Singapore banks embrace AI amid employee job fears

[SINGAPORE] Artificial intelligence (AI) is no longer a distant prospect in banking. It is reshaping the sector, from call centres to analytics to compliance.

Singapore lenders are betting big on the new technology to secure an edge, but a critical challenge remains: ensuring that their staff are not left behind in an AI-driven future.

It is no surprise that employees are uneasy. A Bloomberg Intelligence report in January forecast that global banks could cut as many as 200,000 jobs in the next three to five years, as AI encroaches on tasks currently carried out by humans.

Closer to home, DBS in February said it expects to reduce its contract and temporary staff across its 19 markets by around 4,000 over the next three years, with increased AI adoption.

Even so, banks polled by The Business Times say that the technology will more likely augment rather than replace most roles, provided staff can wield AI as a tool to improve efficiency.

With more than just efficiency at stake, how banks navigate this shift will not only redefine careers in the sector, but also the future of finance.

How banking jobs might shift

The most exposed jobs are those that involve repetitive and predictable tasks.

This includes data entry clerks, customer service representatives, and certain teller functions, all of which are increasingly supported or replaced by digital tools such as chatbots and virtual assistants, says Linda Teo, country manager of ManpowerGroup Singapore.

Entry-level analysts may see parts of their work, such as financial data analysis, supported or replaced by AI-generated insights, she notes.

The technology is also increasingly used for know-your-customer (KYC) processes and transaction monitoring – scanning data for potential fraud, says Randstad Singapore general manager Lim Chai Leng.

As a result, the bar is rising for many traditional entry-level and back-office roles, given that they consist largely of repetitive tasks like data entry and document verification, she says.

“This has shifted hiring requirements towards talent who can work alongside technology to analyse data, improve customer experience, and guide financial decision-making,” Lim adds.

This will change the shape of banking careers, with less entry level roles available, says Gaurav Kwatra, partner and head of Quotient, AI by Oliver Wyman, Asia Pacific.

“But organisations will also try to balance it and keep the future in mind, because they want to have the future senior bench in place as well,” he says.

ManpowerGroup’s Teo also expects the demand for temporary and contract roles will start to decline, especially for those that do not require specialist skills.

These roles have traditionally been used to supplement the core workforce during seasonal surges or short-term projects, she notes.

She says: “Tasks that were previously labour-intensive or time sensitive can now be handled more consistently and cost effectively through automation.”

Augment, not replace

Nevertheless, banks that BT spoke to are less certain that AI can replace jobs so easily, at least for now.

“With all technologies, there’s always talk that is going to eliminate or decimate lots of roles, but history has proven us all wrong time and time again,” says Yap Aye Wee, OCBC’s head of learning and transformation for group human resources.

“What really happens is that when certain roles get eliminated, new ones emerge.”

When DBS said it would reduce its contract or temporary workforce, it also announced it would add 1,000 roles that are emerging due to AI.

Meanwhile, UOB is hiring staff with AI experience to enable its AI capabilities as well, says head of group human resources Dean Tong.

Tong notes that staff numbers remain steady, but points out that UOB – which used to grow its headcount by 1,000 a year – has been able to keep its headcount constant over the last two years.

“You can say that we are now doing more with the same (number of staff) with the help of technology,” he says.

There also exist several roles that are unlikely to be fully replaced by AI, experts say.

Positions requiring emotional intelligence, complex problem-solving and human connection – such as compliance officers, regulatory advisers, relationship managers and investment bankers – may be less vulnerable to automation. But even these jobs are being reshaped by technology.

For example, Dorothy Neo, customer service manager at OCBC’s contact centre, says the bank’s speech-to-text AI tool has allowed her to have more time to analyse calls and provide insights to improve customer service.

The 28-year-old notes that team managers previously spent a lot of time listening to and transcribing calls manually, but the AI tool can now transcribe calls within three to five minutes, freeing up time for deeper analysis.

Meanwhile, Jasmine Lim (not her real name), who works as a corporate banking relationship manager at a local bank, says her company’s internal co-pilot helps her summarise meetings and quickly find documents buried in its intranet, says the 50-year-old.

At UOB, Gen AI has been embedded into the bank’s call centre to help agents answer questions, Tong says.

In the future, he expects a machine to answer simpler calls, while more complicated calls are routed to a human being.

This would help alleviate the volume of calls at UOB’s call centres, which have consistently risen in recent years.

“Sometimes you might find that (the volume growth) may not be so sustainable as well, because you cannot find so many call centre agents,” he says.

But OCBC’s Neo believes human oversight remains critical.

“The speech-to-text tool can only capture words, but not emotions, sentiments or tone of the customer. The outcome (of the AI analysis) may be ambiguous, and that is where a human needs to step in to analyse if this is really the outcome,” she says.

Wong Yang-Sheng, Standard Chartered Bank’s head of human resources for Singapore and Asean, agrees: “While Gen AI can augment our employees and improve our productivity and competitive advantage, it cannot replace decision-making and judgement.”

AI skills training

To prepare their employees for AI, most banks train them in basic AI literacy and provide more in-depth training for specific departments.

DBS says 12,000 employees have begun upskilling or reskilling this year. These skills include AI and data, as well as business domain, change management, soft skills and on-the-job training.

“Over the years, thousands of employees have upskilled and taken on new roles in the bank,” says a DBS spokesperson.

“More recently, we doubled down on our upskilling efforts to ensure that employees continue to stay relevant even as Gen AI reshapes operating models.”

OCBC’s Yap says the bank’s AI training is role-specific and just-in-time, so that it remains relevant for the staff. For those who do not really use much AI in their day-to-day work, the bank also provides foundational training, such as how to identify deep fakes.

Within the organisation, Yap notes that the skills of leaders would have to change as well.

“Traditionally, leaders are either very good at performing or very good at transforming, but increasingly we need to find ways to inject that dual capability in our leaders,” she adds.

Banks are also growing their own AI talent internally.

For example, UOB’s AI Data Analytics Centre of Excellence takes in graduates for a two-year programme that provides certifiable AI skills training and job placement opportunities.

It also has a Technology Development Programme, which trains new hires with non-tech backgrounds into technology talent.

Meanwhile, digital banks such as GXS Bank say their teams and roles were designed to benefit from technologies such as AI.

“We are in a unique position to tap the potential of AI because we are a bank that is built from scratch and therefore do not have any legacy limitations,” says GXS Bank group head of people Amy Tan.

She says the bank regularly reviews its roles to ensure that people are ready to take on new opportunities brought about by AI.

Soft skills

The pervasiveness of AI also means that employers are increasingly looking at soft skills when evaluating new hires.

Randstad’s Lim says employers are now using behavioural interview questions and skills assessments to evaluate a candidate’s true understanding of applications.

UOB’s Tong says: “Gen AI will be a commodity – every bank will be using it, so the differentiator is not in using it, but in creating trust.”

“For that reason, we have doubled down on the screening for people who meet (the bank’s values).”

For example, a candidate may be allowed to use any tool to work on a take-home assignment, and interviewers will evaluate them on how they present their solutions, he says.

“If you completely relied on Gen AI to produce the solution for you and have not internalised it, you will not be able to answer some questions during the interview, compared with someone who has leveraged the tool but augmented it with their own thinking,” he says.

OCBC’s Yap stresses the need for critical thinking.

She says: “A new entrant today is going to have an experience that’s quite different from a new entrant from even five years ago, because they will have that at their disposal a range of AI and generative AI tools.”

“The approach that we’re advocating is that our new entrants look at these AI tools as tools, not really as a master, but more as a servant.”

Yap encourages new hires to pick a human mentor, as the apprenticeship is still important to help understand nuances and complexities which may not be captured by AI.

Why are banks using AI?

Amid a competitive landscape, banks are looking to leverage emerging technologies such as AI to re-think their operations and offerings to deliver better service, said Mohit Mehrotra, financial services industry leader for Deloitte South-east Asia.

Deloitte estimates that banks that successfully innovate with AI can achieve a 5 to 15 per cent improvement in cost-to-income ratio over the next five years.

For example, in Citi’s technology department, over 30,000 developers use AI to automate more than 740,000 code reviews, says Arun Mirchandani, MD of Asia South cluster and banking head of human resources.

This has already saved an estimated 100,000 hours a week, time they now reinvest directly into innovation, Mirchandani says.

DBS also notes that their Customer Service Officer (CSO) Assistant has reduced call handling time by up to 20 per cent, with transcription and solutioning accuracy of over 95 per cent.

The CSO Assistant helps CSOs transcribe customer queries in real-time and perform “live” searches on the bank’s knowledge base to quickly retrieve information, says the DBS spokesperson.

Singapore’s banks are already among the most advanced in the region when it comes to adopting AI, says Oliver Wyman’s Kwatra.

A supportive regulator, strong digital infrastructure, concentrated tech talent and active fintech community mean Singapore is likely to be one of the first places where new AI capabilities are tested and scaled, he says.

Most banks that BT spoke to had developed their own internal co-pilot tools, such as OCBC GPT, DBS-GPT, SC GPT and Citi AI.

Kwatra notes that banks have a lot of data and talent, and have built capabilities around machine learning and analytics in the past, putting them in a good position to adapt to AI.

“What slowed (them) down… is that the banks are also regulated. There are additional fiduciary responsibilities, which, given the stochastic nature of AI, you want to be careful in rolling it out,” he says.

He expects banks that can develop methodologies around their checks and balances are able to move faster in their AI journey.

Nevertheless, UOB’s Tong expects there is more advantage to being a fast follower, than the first mover.

“Because we’re dealing with regulators, customers’ information, and employees’ information, you have to be very careful,” Tong says.

“The first one that does this will probably take a disproportionate amount of risk to clear the road for the rest to come in.”

Looking ahead, modern banks are no longer managing a single workforce but five interdependent ones, says Nesan Govender, talent and organisation lead for South-east Asia at Accenture.

These are human-led roles, human-machine collaboration, robotic-driven automation, Gen AI, and fully autonomous agentic AI.

“The lesson is clear: AI is only valuable when paired with a workforce that can wield it effectively,” Govender says.

“The banks thriving are those treating AI as an enhancer, not a replacer.”