The question for Singapore’s boardrooms is no longer whether to adopt AI. That decision has already been made. The question is whether adoption and governance are scaling at the same pace. In many organisations, they are not.
There are two sources of competitive advantage in the AI economy. The first is productivity: the gains from deploying AI at scale. The second is less recognised, but increasingly material — the governance premium that accrues to institutions able to deploy AI responsibly. Both generate returns. Increasingly, they reinforce each other.
The first alpha: what AI adoption delivers
The case for AI productivity and efficiency is clear. Across financial services, healthcare, law and logistics, AI compresses decision cycles, surfaces patterns in data that human analysts would miss, and frees skilled professionals from work that does not require judgment. The dividend is tangible and measurable in cost and time savings, and increased revenue.
See also: OpenAI is so yesterday — even for SoftBank
Singapore businesses have been early and active adopters of AI, to the nation’s advantage. The risk, however, is treating AI deployment as an end in itself.
DBS CEO Tan Su Shan has highlighted that AI governance remains paramount, noting that the bank’s responsible AI framework provides “a firm foundation in understanding how best to deploy AI agents safely”.
That framing matters: not governance as a constraint on adoption, but as the condition that makes adoption sustainable and scalable. This is a distinction many organisations have yet to grasp.
See also: Anthropic warns investors to avoid certain secondary market sellers
Evidence suggests AI adoption remains uneven. McKinsey’s State of AI 2025 report finds adoption concentrated in front-office functions such as sales and marketing, while legal, risk and compliance lag behind. This means those best placed to govern AI are often least familiar with it. This gap, if uncorrected, will lead to undesirable consequences.
Respondents who said “don’t know” or “other” are not shown. In media and telecom, n = 98; insurance, n = 61; technology, n = 249; healthcare, n = 101; consumer goods and retail, n = 129; professional services, n = 291; travel, logistics, and infrastructure, n = 66; energy and materials, n = 191; banking and other financial institutions, n = 152; advanced manufacturing (includes advanced electronics, aerospace, automotive and assembly, and semiconductors), n = 136; engineering, construction, and building materials, n = 90; pharmaceuticals and medical products, n = 77.
Source: McKinsey Global Survey on the state of AI, 1,993 participants at all levels of the organization, June 25–July 29, 2025
When AI failure is a governance failure
In 2023, the US’s largest healthcare company, UnitedHealth Group, through its subsidiary naviHealth, deployed an AI tool called “nH Predict” to estimate appropriate lengths of post-acute care.
UnitedHealth is now being sued in court, where plaintiffs allege that the AI system produced unfair and incorrect recommendations, with an error rate as high as 90%, raising questions about its reliability as a substitute for clinical judgment. The case is ongoing, and in March, a US federal court ordered discovery into the algorithm’s use, signalling that its AI deployment is now under substantive judicial scrutiny.
Sink your teeth into in-depth insights from our contributors, and dive into financial and economic trends
The failure here was not that the algorithm malfunctioned. It operated as designed. The failure was that no effective governance mechanism existed to challenge outcomes, escalate concerns, or assign clear accountability when risks emerged.
When courts and regulators scrutinise algorithmic decision-making, the question is rarely how the model worked. It is who approved it, who monitored it, and who was responsible for its outcomes. That principle applies in Singapore, where regulators are increasingly emphasising accountable AI, governance oversight and responsible deployment.
From failure to governance
Two principles follow for boards and management teams.
First, accountability cannot be outsourced. Delegating a decision to an algorithm does not transfer liability for its outcomes. Nor does reliance on a third-party vendor insulate the deploying organisation.
The governance obligation — to understand what a system does, to whom, and with what consequences — remains with the institution that chose to use it.
Second, silence is a decision. Regulatory frameworks — including the EU AI Act, which has extraterritorial reach for companies operating in or supplying European markets — are increasingly treating awareness of risk without remedial action as grounds for liability. The question is not whether AI is used, but whether those responsible for governance have the information, authority, and mandate to act when something is wrong.
The second alpha: why governance generates returns
Corporate governance is conventionally treated as a cost of doing business - a function that exists to satisfy regulators and avoid penalties. The more interesting argument is that it generates returns.
On Feb 12, GIC led Anthropic’s US$30 billion ($38 billion) Series G funding round at a post-money valuation of US$380 billion, building on its earlier participation in the company’s Series F in September 2025. Temasek co-invested in the Series G.
Fifteen days later, the US Pentagon designated Anthropic a supply chain risk. The reason: Anthropic had refused to allow the Pentagon to deploy its AI model for mass domestic surveillance of US citizens, or to power fully autonomous weapons systems.
While GIC and Temasek did not invest with knowledge of what followed, they invested in a company whose governance position then created an unexpected confrontation with the world’s most powerful government - and neither has retreated from that investment.
The point here is that GIC and Temasek, as sophisticated long-term investors, are prepared to back companies that hold governance positions under pressure. In sectors where AI will increasingly operate, for example, finance, healthcare, law and public infrastructure, that calculus is sound. A premium is ever more being placed on trust and security. Regulators extend latitude to institutions they regard as trustworthy. Enterprise clients increasingly make procurement decisions influenced by governance considerations. Investors model reputational and regulatory exposure.
The organisation that demonstrates coherent AI oversight signals maturity, care, and trust. It is also building a moat against competitors who, without that governance infrastructure, will find it difficult to replicate trust quickly.
Turning governance into the second alpha
Singapore regulators are increasingly explicit that AI governance sits with senior leadership.
In May 2026, the Cyber Security Agency of Singapore urged all Critical Information Infrastructure operators to review cybersecurity readiness in light of accelerating AI-driven threats, noting the narrowing window between vulnerability discovery and exploitation. In Parliament the same day, it was emphasised that AI governance is the responsibility of boards, not a matter for IT departments alone.
But regulation alone does not create advantage. Governance becomes the second alpha only when responsibility for AI is clearly assigned, operationalised, and embedded in management accountability.
These crystallise into three governance questions:
• Are AI risks embedded in your enterprise risk framework alongside cyber and operational risk?
• Do you maintain a complete, continuously updated inventory of AI systems, with named accountability for each deployment?
• If an AI system were producing harmful outcomes today, how quickly would you detect it, and who would be responsible for escalation?
How organisations answer these questions will determine whether governance becomes a cost centre or a second alpha.
When the two alphas compound
What has changed is not the underlying governance issues of accountability, trust, and control. It is the scale and speed at which they now manifest.
Governance is not the opposite of growth. It is a condition of it. The organisations that understand both alphas — and pursue them together — will find they compound.
Kay Pang is a Singapore board director and business lawyer specialising in technology law and AI governance. She has advised boards and senior management on AI adoption and governance, contributed to World Economic Forum white papers on the responsible use of technology, and speaks regularly on AI law and ethics.
