Artificial intelligence and the future of financial work
Read the PDF edition (3 pages)
A workforce of 200,000
The financial sector is one of Singapore’s largest employers of skilled labour. In May 2026 Deputy Prime Minister Gan Kim Yong, who also chairs MAS, put its contribution at about fourteen per cent of gross domestic product and its workforce at around 200,000 professionals, more than eighty per cent of them locals.1 The share of local employees matters politically. Part J describes how Singapore balances the attraction of foreign talent against the expectations of its citizens. A technology that displaces local staff in back-office and middle-office roles would sharpen this tension.
The tasks most exposed to artificial intelligence are those with repetitive and documentary content: processing operations, checking and verifying data, producing standard reports, the first review of compliance alerts and the drafting of routine correspondence. Tasks that combine relationships, judgment and responsibility are less exposed, among them advising complex clients, negotiating, taking credit decisions and dealing with regulators. Between these poles, most roles are likely to shift towards supervising tools, interpreting their output and handling the cases that tools cannot resolve. The examples that MAS and IBF chose to present in 2026 illustrate this shift. A customer service officer now leads a team that develops artificial intelligence for customer experience, while a frontline team leader helps to oversee the bank’s customer chatbot. A UOB employee moved after thirty-seven years in wholesale banking into a retail anti-money laundering role through the bank’s reskilling programme.2
Tripartite retraining
Singapore’s response follows the tripartite model of government, employers and unions that has shaped its labour market since independence. In September 2026 the government reported that IBF, the financial and professional services cluster of the National Trades Union Congress and seven industry associations were enhancing their memorandum of understanding. Together they represent more than 800 financial institutions and fintech firms and over seventy per cent of the sector’s workforce.3
The centrepiece is the IBF AI Workforce Co-Lab, which brought together twenty-three pioneer institutions from banking, insurance and asset management. They committed to train their entire Singapore workforce, more than 80,000 employees, in critical artificial intelligence skills by 2028, of whom more than half had been trained by September 2026. The Co-Lab has three pathways: one for leaders on strategy, governance and workforce transformation, one for wealth managers that combines AI skills with advisory expertise, and one that redesigns operational roles and career paths. Ten private banking employers committed to upskill their relationship managers.4 By the same date there were nearly sixty IBF-accredited programmes teaching AI skills to finance professionals, contextualised to financial-sector use cases and MAS guidelines.5
The political message was explicit. Financial institutions, the Deputy Prime Minister said, should “plan for their people even as they plan for AI”. The government wanted artificial intelligence “to translate into better careers for Singaporeans”, he added.6 The statement reflects a familiar bargain. The state supports adoption and funds retraining. In return it expects employers to redeploy staff where they can. Whether the bargain holds when automation accelerates will be one of the tests of the coming decade, which Chapter K.VI considers in its scenarios.
New entrants and the skills that count
The pipeline of new entrants is changing as well. In May 2026 IBF announced a Young Talent Programme for AI in Finance, developed with the Singapore Management University, Nanyang Technological University, financial institutions and technology partners. Students work on industry projects, receive mentoring from practitioners and may earn academic credit. Strong performers can obtain AI-related internships or traineeships. The pilot started in August 2026 with around twenty financial institutions. Gan’s summary of the skills required was precise: “Workers who are comfortable using AI, and who also have strong domain knowledge, will have a clear advantage.”7
The emphasis on domain knowledge deserves attention. Artificial intelligence makes it easy to produce a plausible analysis of a fund structure, a sanctions question or a tax position. Only someone who understands the subject can tell whether the analysis is right. The financial centre therefore needs professionals with deep knowledge of law, tax, credit and markets more than ever, because they are the ones able to verify the output of the tools. The same point applies to language, which Part J treats as a working asset of the financial centre.
Worked example
A Singapore bank’s compliance unit handles 15,000 transaction monitoring alerts a year. About ninety-five per cent are closed as unfounded after an average review of thirty minutes, which amounts to roughly 7,100 hours. The bank deploys a machine-learning model that ranks alerts and proposes closure for those with the lowest risk scores, with an analyst confirming each proposal. Under the Guidelines, the bank would in all likelihood assess the use case as of high risk materiality, because failure would affect its compliance with anti-money laundering law. The bank therefore records the model in its AI inventory, has it validated by a team independent of its developers, documents its performance against historical cases and sets thresholds above which cases are always escalated. If review time per alert falls by half, the unit saves around 3,500 hours and redeploys them to complex cases and periodic reviews of high-risk relationships. The senior manager responsible for financial crime compliance remains answerable for the outcome.
Notes
- Gan Kim Yong, keynote speech at the IBF Financial Industry Fiesta, 19 May 2026 (mas.gov.sg): the financial sector contributes about 14 per cent of GDP and employs around 200,000 professionals, “more than 80%” of them locals. ↩
- Gan Kim Yong, keynote address at the IBF Distinction Evening, 24 September 2026 (mas.gov.sg): examples of DBS and UOB employees who moved into roles overseeing or developing AI, including a UOB employee who moved after 37 years, through the Better U Pivot programme in 2025, from wholesale banking into a retail anti-money laundering role. ↩
- Gan Kim Yong, keynote address at the IBF Distinction Evening, 24 September 2026 (above): enhanced tripartite MOU between IBF, the NTUC Financial and Professional Services Cluster of Unions and seven industry associations, representing over 800 financial institutions and FinTech firms and more than 70 per cent of the sector’s workforce. ↩
- Gan Kim Yong, keynote address at the IBF Distinction Evening, 24 September 2026 (above): 23 pioneer institutions; over 80,000 employees to be trained by 2028, more than half already trained; Co-Lab@Leaders, Co-Lab@Wealth Managers and Co-Lab@Operations; 10 employers committed under Co-Lab@Wealth Managers; Job Redesign Playbook for Financial Services. ↩
- Chia Der Jiun, Managing Director of MAS, welcome remarks at the IBF Distinction Evening, 24 September 2026 (mas.gov.sg), para 4: “IBF has accredited close to 60 programmes”, developed with training partners and financial institutions, to build critical AI skills and contextualised to financial-sector use cases and MAS guidelines. ↩
- Gan Kim Yong, keynote address at the IBF Distinction Evening, 24 September 2026 (above). ↩
- Gan Kim Yong, keynote speech at the IBF Financial Industry Fiesta, 19 May 2026 (above): YTP-AIF co-developed by IBF, SMU, NTU, financial institutions and technology partners; pilot starting in August with around 20 financial institutions; academic credit from SMU and NTU – Gan Kim Yong, keynote speech at the IBF Financial Industry Fiesta, 19 May 2026, para 19(c). ↩
For general information only. This article does not constitute legal or tax advice, does not replace advice on an individual case and creates no lawyer-client relationship. It reflects the legal position as of October 2026.