According to Bloomberg Business, Singapore’s government, working with the Monetary Authority of Singapore (MAS), is directing the country’s three largest banks—DBS, OCBC, and UOB—to retrain all 35,000 of their domestic staff on AI within the next one to two years. The initiative, involving direct regulator approval for AI tools and a 90% salary support scheme for reskilling, aims to improve skills and prevent mass layoffs as agentic AI automates complex tasks. At OCBC, a tool built by Kelvin Chiang’s team now does in 10 minutes what took private bankers a full day, while DBS uses an internal AI assistant handling over one million prompts monthly. Despite this, DBS still expects to reduce about 4,000 temporary roles in three years, and a Bloomberg Intelligence analysis forecasts the bank could see AI-derived cost savings boosting pretax profit by up to S$1.6 billion.
The Government Hand-in-Glove Approach
Here’s the thing: Singapore’s strategy is uniquely interventionist. It’s not just banks deciding to train people; it’s a coordinated national effort where the regulator, MAS, is literally sitting down with engineers like Kelvin Chiang to vet AI safeguards before rollout. The National Jobs Council and the Institute of Banking and Finance are key partners, even offering to pay banks to map out new career paths for staff. The unspoken driver is pure fear—fear of the “aggressive job cuts” seen in the US and Europe at firms like Goldman Sachs. The government’s bet is that by getting ahead of the disruption with heavy support, they can morph tellers into bankers and compliance officers into AI supervisors. But is that enough?
The Human Reality: Rising Pressure and Quiet Attrition
Look at the workers’ stories. For David, the relationship manager, AI cut his document prep from an hour to 12 minutes. Great, right? But his bosses’ expectations have now skyrocketed, with consultants suggesting each manager could handle 50 clients jumping to 60 or 70. That’s more stress, not less. For Woon Leng, a branch manager in her 60s, AI training is just another pile of homework after an already demanding day. And then there’s the quiet part out loud: economists note banks can pursue headcount reduction simply through “natural attrition”—not replacing people who leave. So, the official line from OCBC and UOB is “no AI-related job cuts,” but the total number of roles can still shrink. That feels like a semantic distinction to the workforce.
The Big Picture: Efficiency vs. Employment
So, is this a humane transition or just a slower, more politically palatable downsizing? The efficiency gains are undeniable and massive. OCBC’s 400 AI models make six million decisions daily. DBS has a tool cutting call times by 20%. The profit boosts are forecasted to be enormous. As one professor bluntly put it, banks would be “completely stupid” to hire now only to lay off in three years. Inaction, as UOB’s AI head said, is a “slow path to irrelevance.” The core tension is that the technology’s potential is so vast we “haven’t really scratched the surface yet.” Retraining 35,000 people is a monumental effort, but if the AI can eventually do the job of 40,000, the math gets ugly. The workforce transformation challenge is global, but Singapore is testing if a top-down, government-backed plan can soften the blow better than the market-driven chaos elsewhere.
The New Competitive Landscape
And what about the next generation? Vania, the 22-year-old AI intern, points out a brutal truth: basic AI literacy is becoming table stakes. “It’s no longer so much of a competitive advantage,” she says. When everyone is getting upskilled, the goalposts move. The real advantage shifts to those who build and govern these systems—the data scientists like Kenneth Zhu, whose team exploded from a handful to over 100. For the average banker, the training might save their job today, but it also resets the baseline. The boss will always want more productivity. The real question for Singapore’s experiment is whether it creates sustainable, higher-value roles for those 35,000 people, or just manages their exit gracefully. My hunch? We’ll see a mix of both.
