Inside Singapore's AI bootcamp retraining 35,000 bankers

[SINGAPORE] Kelvin Chiang knew that the five agent AI models his team built could do in 10 minutes what would take a private banker an entire day. With that in place, he went to Singapore's banking regulator to demonstrate that the safeguards were sufficient to control risk.

Before deploying a tool to create documents for relationship managers in OCBC Bank's private wealth division, he took a team of data scientists less than a mile across the city's downtown financial district to the Monetary Authority of Singapore (MAS). Mr. Chen highlighted the steps banks take when something goes wrong with the blueprint and how bank employees react when the system becomes hallucinatory.

Mr Cheng, 52, head of financial crime compliance analysis at Bank of Singapore, said he felt the regulator's engineers were receptive to the policy after the presentation and discussion with MAS representatives. “They were very happy.”

Singapore's AI efforts highlight a unique balancing act. The government is working with the country's largest banks to teach new technology skills. The implicit intention is to curb the large-scale layoffs seen at some financial institutions in the United States and Europe. The futures of thousands of bankers are at stake as industry giants such as Goldman Sachs Group Inc. tell their staff to expect further job cuts as they leverage AI.

Companies around the world are turning to agent AI because it is more powerful than previous technology iterations, allowing models to perform actions independently and perform multiple steps at once. Singapore is hoping for higher levels of government involvement to pave the way for workers, with a focus on improving skill sets and mitigating the need for mass layoffs.

Singapore's National Development Minister Chee Hong Tat, who is also vice-chairman of MAS, said the three largest banks – DBS, OCBC and UOB – will retrain all 35,000 of their domestic employees over the next one to two years.

“Governments are taking steps because they recognize that this ability and this change actually potentially causes a lot of fear,” said Violet Chan, a senior partner at McKinsey. “Given what we've seen in other markets like the US where we're seeing more aggressive layoffs and layoffs, the government is essentially recognizing that we need to do something as a country for these big companies.”

For David, 39, who manages money for his bank's wealthy clients, smarter technology is helping him get to work faster than at any time in his 16-year career, but it's raising his boss's expectations and making him nervous. Typically, it used to take about an hour to create a single customer order form, but now it only takes 10 to 12 minutes. That means more time in front of customers, he says. He declined to give his full name regarding private details about his work.

Banks will be able to increase the number of clients each relationship manager can serve from, say, 50 to 60 or 70, said Mohan Jayaraman, a senior partner at Bain.

“The argument is that if you can spend more time with your customers, you'll get better coverage,” he says. “You're increasing the interest of the people you're offering and at the same time giving RMs the ability to earn more revenue from a larger pool.”

Singapore's National Employment Council was born out of the country's employment needs during the COVID-19 pandemic and is now also working with the Institute of Not-for-Profit Banking and Finance as a key partner in promoting job growth in the financial industry. In some cases, in response to role changes, the institute will work with financial companies and disburse funds to them to plan how staff can take on redesigned roles or pivot to new career paths. In some scenarios, the company is offering banks up to 90% salary subsidies to reskill mid-level and entry-level employees, according to its website.

MAS says it is working closely with IBF, financial institutions and trade unions to prepare the industry's workforce for AI adoption.

OCBC and UOB said there will be no AI-related job cuts. DBS says it will not lay off any full-time employees, but expects to reduce the number of temporary and contract workers by about 4,000 over the next three years as vendor contracts end. Bloomberg Intelligence analysis predicts that DBS will realize greater cost savings from AI than its peers, boosting pre-tax profits by up to S$1.6 billion (about 17%) over the next few years.

DBS staff currently use an in-house AI assistant that processes more than 1 million prompts per month. We've also built role-specific tools, including tools for customer service representatives, which have reduced call response times by up to 20%. UOB gave all employees access to Microsoft Copilot, deploying over 300 AI use cases across the bank.

Watching from the inside Vania Lim, 22, spent last summer as an AI intern on HSBC Holdings' global payments team. She is optimistic about new roles being created in digital assets and improved wallet use cases, but worries that some of the skills of junior employees are becoming less unique.

“Everyone is now familiar with AI,” she says. “It's no longer a competitive advantage.”

Meanwhile, older staff like Oong Leng, who is in her 60s, say training for new technology feels like an addition to the already demanding job of managing a branch.

She has worked at a local bank for over 40 years and currently uses ChatGPT in her branch to answer customer questions about products. He added that staff are required to take an online course on AI after work and pass a test afterwards, as they are too busy during the day. She declined to give her employer's full name.

Walter Seera, associate professor of economics at the Singapore University of Social Sciences, said many banks choose not to play a role when employees move in and out of the company.

“Of course, companies don't like retrenchment or layoffs, especially when it comes to large companies that are very closely watched by the Singapore economy and the government,” he said. “However, companies can pursue layoffs by simply not filling positions when natural attrition occurs.”

At OCBC, the AI ​​Lab had fewer than a few people when Kenneth Zhu joined the company in 2018. Eight years later, the 36-year-old executive director of data science and AI oversees a lab in the data office that employs more than 100 people. Nearly 400 models make 6 million decisions every day, from flagging suspicious transactions to scoring credit risks to filtering out false positives in anti-money laundering systems that once ate up compliance officers' time.

It's too early to assess whether Singapore's approach will benefit banks and their staff, but Jochen Wirtz, associate dean of the MBA program and professor of marketing at the National University of Singapore, says the technology's potential remains huge and warns companies against holding back.

“My hunch is that we haven't even scratched the surface,” he says. “It would be completely foolish for banks to have a bunch of employees now that they're going to have to lay off again in three or four years. They're much better off freezing now and trying to reskill them as best they can.”

At UOB, we are hard at work building AI fluency across the organization, as well as training our staff, including through our internal 'Better U Pivot' programme.

“Doing nothing isn't really an option for us or anyone,” said Alvin Eng, head of enterprise AI at the company. “It’s a slow road to irrelevance.”Bloomberg