Reinventing finance in the digital age

OVER the past 15 years or so, the rise of digitalisation and fintech has been one of the most significant forces that has reshaped the financial landscape globally and in Singapore. It has shaped who is in the financial system, how finance flows, and what risks are present.

A significant development in the industry is undoubtedly the use of artificial intelligence (AI). Financial institutions now leverage AI and data to better understand consumers and more effectively deliver personalised experiences, products and services in real time. Functionally, AI has enabled non-traditional credit underwriting, allowing financial institutions to take a more nuanced approach to credit provisioning. It has also improved investment options.

Specific AI-powered tools are already benefiting customers. For example, in Singapore, DBS customers who used the bank’s AI-powered financial planning tool saved twice as much, invested five times more, and were nearly three times more insured than those who did not engage with the tool.

OCBC provides an AI-powered stock-maker tool that predicts stock-buyer movements and generates hyper-personalised stock ideas from that. UOB Asset Management’s AI-augmented investment management process integrates AI and machine learning techniques with analyst research capabilities to produce better investment outcomes.

AI adoption is now widespread across various banking functions. When I retired from DBS, the bank was using some 1,500 AI models across almost 400 different business cases – in risk management, financial management, marketing, audit and compliance. All this has changed the way the system functions.

Impact on financial roles and jobs

The growing integration of AI poses direct questions for the industry’s employers and employees. A pertinent one is: How do you maintain morale and productivity when employees know that AI might automate their jobs?

The challenge is severe because the integration of AI use within the office environment is becoming more pronounced. White-collar jobs typically involve four activities: reading, synthesising, producing output and taking action.

AI is increasingly capable across all these areas, often faring better than humans. It even produces presentations and output. And it also does accounting and distribution.

So it is not illogical to believe that in the short term, AI could actually wind up displacing a lot of the jobs as we know them.

SEE ALSO

Singapore’s ‘secret sauce’ is balancing trust, stability and unorthodox innovation: Piyush Gupta

SFF 2025: How AI, tokenisation and quantum can power the next decade of fintech growth

Stopping the conveyor belt of crime

Transition challenges and future skills

While the transition will not be easy and there will be job loss, I believe society will eventually find ways to adapt. To manage this shift, we need to work seriously at reskilling, teaching people skills and equipping them with tools to move into the future.

Successful reskilling is possible. In 2016, when DBS started digitalisation, we identified 1,200 jobs for automation as well as engaged the unions and staff. We reskilled all workers involved and found corresponding jobs – 200 left and retired while 1,000 transitioned. With technology displacement, new jobs are also created.

The nature of skills required is changing rapidly. Five years ago, coding was identified as the area with the largest skills shortage. In the last two years, all tech companies have stopped hiring coders. The era of domain specialisation is fading. Instead, the future will belong to people who can establish context, look to the horizon, find solutions and piece things together.

Ethical and regulatory considerations

Looking beyond the banking system itself, the improved ability of AI to discriminate also raises ethical issues for society. When we build an AI model that can discriminate better than humans can, it could lead to red-lining, or denying services to certain population segments based on data and analytics.

So how should society balance between achieving efficiency and other goals, like helping the disadvantaged?

To address the potential misuse of AI and data, the Monetary Authority of Singapore (MAS) and the industry have established regulatory frameworks, such as the Fairness, Ethics, Accountability, and Transparency principles, which offer firms providing financial solutions a set of principles to consider when using AI in data.

MAS has also introduced the Artificial Intelligence Model Risk Management guidelines with three focus areas covering AI oversight and governance, key risk management systems and processes, as well as the development and deployment of AI.

And there is Project Mindforge, driven by Veritas Initiative and supported by a consortium comprising MAS, Accenture, financial institutions and technology partners, which aims to develop a framework for responsible use of generative AI in the financial services industry. The goal is to encourage power innovation in solving common industry-wide challenges and enhance risk management.

These initiatives represent the critical, collaborative efforts needed to build the ethical guardrails for this new technology. Close partnership is what ensures that as we reinvent finance with AI, we do so in a way that builds trust and serves society responsibly.

The writer is the 17th S R Nathan Fellow at the Institute of Policy Studies (IPS). He was the former chief executive officer and director of DBS Group.

This is an edited excerpt of the writer’s second IPS-Nathan Lecture and question-and-answer session.