How Singapore’s business leaders are positioning AI governance: Opinion

Artificial intelligence is developing fast, yet its governance in businesses is barely catching up. Often, it is like fixing a plane while flying it.

Business leaders will be familiar with it. The shock first came from the emergence of generative AI (GenAI), which has been a wake-up call to workers, with its speed and efficiency at day-to-day tasks.

Agentic AI is an even greater game changer, because it can independently make decisions and take actions. Coming next is artificial general intelligence, which will have human-like reasoning and cognitive powers.

The fundamental concern is how AI can be postured, or in fact, governed.

Return, risk and responsibility

Drawing from an exploratory analysis of large corporations, as well as small and medium-sized enterprises (SMEs), I propose three perspectives on how businesses can position AI governance: return, risk and responsibility.

A natural motivation for companies to adopt AI is for the business benefits or return.

AI models may also contain biases, which can lead to discrimination and unfair practices. Other risks are systemic, such as the loss of human control over AI or overdependency on the technology, which can affect the thinking of business leaders.

From the responsibility angle, AI governance will have to address privacy, safety and security, as well as human accountability.

In addition, there are broader social aspects to AI use that companies will have to tackle, such as the displacement and retrenchment of workers.

Businesses have begun to progressively practise and report on AI governance, but the scale differs.

Larger AI adopters, such as DBS and Keppel, have been more holistic in managing all three perspectives. SMEs, on the other hand, are mostly beginning the journey and are focused almost solely on deriving business return.

Lessons from DBS

The largest bank in Singapore and South-east Asia by total assets, DBS was named “World’s Best AI Bank” by the Global Finance periodical in its inaugural AI In Finance Awards in 2025.

DBS describes itself as an “AI-fuelled bank” and excels in leveraging return from the technology.

Upon winning the award, the bank stated that it has implemented over 1,500 AI models in more than 370 use cases, with a projected economic impact exceeding S$1 billion in 2025.

An early adopter of AI, DBS has used this technology to provide SMEs with pre-emptive alerts of credit risks, even before any problem emerges.

This has saved many of these borrowers from default and hence strengthened the bank’s clientele base.

DBS deals with the risk aspect of AI in many ways.

In particular, for high-impact areas such as credit, fraud, collections and customer treatment, it conducts model validation, which involves testing for performance and bias, as well as ongoing monitoring and periodic reviews.

To use AI responsibly, the bank commits to principles including fairness, ethics, accountability and transparency, which are standards promulgated by the Monetary Authority of Singapore.

DBS is also aware that possible job displacement by AI is a critical issue.

Tan Su Shan quipped that, on the day she was appointed as CEO in 2025, the board sent her a WhatsApp message: “You know, even a CEO’s job can be replaced by AI.”

Keppel’s strong AI frameworks

Beginning as a shipyard, Keppel is positioning itself as a global asset manager and operator in dealing with infrastructure, real estate and connectivity.

From a return perspective, Keppel uses GenAI tools to support its investment and research, as well as its data centre operations.

It has deployed Alpha, a proprietary AI tool, to assist investment managers in analysing and evaluating acquisition targets, and rolled out Keppel Data Exchange as a unified source of information for decision-making and reporting.

Keppel’s enterprise risk-management framework ensures that AI adoption is accompanied by enhanced end-to-end cybersecurity monitoring and a zero-trust policy.

This safeguards information technology access, which is granted based on stringent verification and continuous monitoring.

For responsibility, Keppel has a set of internal corporate guidelines called the 10 Commandments for Safe AI Exploration.

It draws on external guidelines such as Singapore’s Model AI Governance Framework and the EU’s AI Act.

The guidelines manage the use of AI by individuals and across projects to uphold data privacy, accountability, ethics and cybersecurity.

SMEs focus on return

However, AI governance is still embryonic for SMEs, which face constraints in resources and capabilities.

Their positioning of AI governance has largely been focused on the return perspective. The technology is deployed to augment the business for financial payback.

For instance, RS Facilities Services, a cleaning sub-contractor, manages more than 1,400 cleaners in over 100 client sites, spanning residential, institutional, industrial and healthcare settings.

Due to its overall digitalisation, it has achieved 99 per cent faster report compilation time, 50 per cent increase in manpower productivity and over 80 per cent rise in hygiene standards.

The company is implementing a novel Toilet Cleaning Robotics solution that will feature an AI-powered stain detection, with cameras assessing cleanliness after each pass and triggering a repeat cycle if the standard is not met.

Another example is floral delivery service Far East Flora, which reportedly grew online sales by 25 per cent and repeat purchases by 15 per cent through the use of integrated digital solutions.

Moreover, AI helped it shave about 50 per cent off the time normally spent on producing marketing materials.

As part of its digitalisation strategy, Far East Flora adopted three AI solutions: namely, a GenAI Digital Training System to deliver mobile microlearning modules for staff; a GenAI Customer Engagement Chatbot to consolidate and reply to enquiries from all channels; and a GenAI Marketing and Sales Content Generation solution for copywriting, image enhancement and campaign management.

Other notable SME AI practices, particularly in GenAI-driven customer engagement and relationship management, straddle various industries.

Companies involved in these practices include Durian Memories – a shop selling durian products – investment advisory Endowus, EU Holidays, Firefly Photography and many others.

However, based on my exploratory study of the SMEs’ publicly available documents on AI governance, there appears to be few formal considerations of the risk and responsibility aspects of using the technology.

The 3Rs in tandem

To govern AI effectively, businesses must go beyond just return.

Risks come next, and then there is the ultimate consideration of responsibility, which has broader ethical and social implications.

Individually and collectively, businesses of all sizes and in all industries cannot have AI governance as a piecemeal afterthought.

The writer is director of the Centre for Governance and Sustainability as well as professor in practice of strategy and policy at National University of Singapore’s Business School

Source: The Business Times © SPH Media Limited. Permission required for reproduction.

ADV: Drafting CPF Division Orders: Getting It Right Freesia Woods condo management ordered to compensate couple over termite infestation