Singapore’s AI momentum is hard to miss. As per the IMDA Singapore Digital Economy Report 2025, three in four workers now use AI at work, reflecting how quickly AI has moved from experimentation to everyday use. Policy is keeping pace, too, with Budget 2026 introducing a National AI Council and a 400% tax deduction on AI expenditure to accelerate enterprise adoption.
At the same time, capital is moving decisively. Temasek Holdings has joined a consortium including Nvidia, BlackRock, and Microsoft in a US$40 billion acquisition of Aligned Data Centers, with ambitions to deploy up to US$100 billion into AI infrastructure. This reinforces Singapore’s position as a global AI infrastructure hub.
On paper, the picture is coherent, with strong policy direction, deep capital alignment, and widespread enterprise adoption. However, access and transformation are not the same. Are enterprises truly transforming how work gets done, or are they simply layering AI on top of existing habits?
Most organisations have moved quickly to enable access, with employees using AI tools, experimenting with prompts, and integrating AI into daily tasks. But few have redefined how work should change because of it. So, while there are faster emails, quicker summaries, and more efficient searches, there is a lack of deep integration of AI into business workflows and decision-making. The gap between usage and transformation is becoming the defining challenge of this next phase. Closing it will depend on whether organisations can make these four structural shifts.
Measure impact, not just adoption
Tools deployed, employees trained, and use cases piloted signal effort, but not transformation. What matters is whether AI shows up in speed, efficiency, customer satisfaction scores, cost per decision, and time to market. If there is no improvement in these productivity metrics, there is no real transformation, just an annual report talking point.
What Singapore’s corporate sector urgently needs is an internal evaluation framework that evaluates model performance, safety risk, and business fit before deployment, and measures real productivity gains post-deployment. Without that, organisations will continue to spend on AI, report adoption rates, but fail to see meaningful improvements in margins or business growth.
Upskilling the right way
A key component of adoption is upskilling. According to a study by General Assembly, 69% of employers in Singapore believe upskilling will play a major role in filling talent gaps in 2026, but companies are balancing training with hiring and outsourcing rather than treating upskilling as a primary solution. This balance is driven by the need for immediate agility versus long-term resilience.
However, it’s important for employers to remember that upskilling should be treated as a core business strategy, as it cultivates institutional knowledge and employee retention, while outsourcing allows firms to bypass time-intensive training and may serve as a short-term solution.
A major downside of outsourcing and hiring is the potential for excessive dependence on external and new resources. An added benefit of upskilling is increased employee engagement and retention, as employees can see a future for themselves within the company.
Develop applied AI capabilities
While the Singapore government and companies in the country have invested heavily in broad AI literacy, an area that needs more attention is domain-specific expertise. The financial analyst who can interrogate model outputs rather than accept them, the logistics engineer who can retrain a routing model, and the clinician who can recognise when a diagnostic algorithm is confidently wrong. These are the capabilities that separate AI adoption from real value creation.
A General Assembly study found that nearly 58% of Singapore employers struggle most to hire data analytics and data science talent, well above the United States at 44% and the United Kingdom at 43%. These are not entry-level gaps. They are the roles that determine whether AI creates value for the business or simply creates noise.
The deeper issue is focus. Learning and development efforts are scaling broadly, but not with sufficient depth, and rarely in the context of specific functions. Training that teaches employees how to use a prompt is not the same as training that enables them to translate business problems into data questions, stress-test outputs, and make better decisions.
Move beyond sponsorship to leadership participation
Singapore leaders are watching, and many of them are asking the right questions. They know AI is changing what their teams need to do. Fewer know how to adapt to an AI-driven workplace. This gap is one of the most overlooked reasons deployments fail.
In many organisations, leaders approve budgets, launch initiatives, and set targets, but often do not use AI in their own work. Employees notice. This sends the message that only employees, not leaders, are expected to use AI in their work. Adoption stalls, and scepticism builds quietly.
What shifts the dynamic is not executives becoming AI experts, but becoming visible users and learners who are willing to accept both the successes and failures of AI implementation in their work. The leaders who move organisations forward are those willing to experiment and share: “I tried this yesterday, and it did not work the way I expected. Here is what I learned from it.” This will do more for AI adoption than any top-down mandate.
Singapore’s AI advantage is real. Whether it remains statistical or becomes structural depends on decisions being made right now inside boardrooms, within L&D priorities, and in how leaders use AI themselves. The shift is not just about adopting AI, but about building the capabilities that make it matter. Organisations that do this well will not just keep pace with change; they will help define it.
