As tariffs bite into margins, AI investments must deliver

Business Times

28-04-2025

ON APR 2, the United States announced a broad-based 10 per cent baseline tariff on all imports, in addition to 'reciprocal tariffs' on a few dozen countries, as part of its latest trade measures. The policy is applied uniformly across trading partners – including longstanding economic partners like Singapore, with whom the US runs a trade surplus.
For Singapore, built on the principles of free trade, this marks a significant inflection point. As Prime Minister Lawrence Wong warned in his recent S Rajaratnam Lecture: 'The global order we took for granted is fraying, and the risks of fragmentation and disorder are growing.' For small, open economies like Singapore, this development poses existential risks.
Yet it's precisely in this moment of global fragility that Singapore must respond with clarity and conviction – not least in the technological and digital infrastructure we've built, the talent we've nurtured, and the transformation model we've pursued.
Keeping faith in AI investments
As artificial-intelligence funding accelerates across the region, the focus is shifting from ambition to execution – where returns on infrastructure, talent and tools must now be proven in real business terms.
Singapore has committed over S$1 billion from 2024 to 2029 under its National AI Strategy 2.0, including up to S$500 million for advanced chips and compute infrastructure, S$150 million for enterprise AI adoption, and S$120 million for AI-driven scientific research. More than S$20 million will also go towards scholarships and training to triple the national AI talent pool to over 15,000 practitioners. These investments must now pay off, in productivity, efficiency and margin resilience.
Across Asia-Pacific, businesses implementing AI and automation are expected to achieve 20 to 30 per cent reductions in operational costs and over 40 per cent gains in efficiency. Deloitte's Global Intelligent Automation survey found that organisations that advanced their automation efforts reported average cost savings of 32 per cent. Forrester highlights that AI-powered automation can reduce operational costs by up to 30 per cent, particularly by optimising workflows and reducing human error. And 72 per cent of data and IT leaders surveyed in the Dataiku 2025 GenAI Trends Report declared positive returns on investment from their generative AI projects.
BT in your inbox
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
Sign Up
Sign Up
Amid the tariff uncertainty, every percentage point in enterprise savings and efficiency counts.
Doubling down on Singapore's AI future
In Singapore, AI-readiness is the result of years of sustained investment, forward-looking policy, and strong public-private collaboration. Global firms such as Amazon, Google and Microsoft are deepening their presence through long-term investments – part of the S$13.5 billion in overall commitments made in 2024 alone, reflecting strong confidence in Singapore's digital infrastructure and talent base.
For instance, Amazon Web Services (AWS) announced a S$12 billion investment in Singapore's cloud infrastructure through 2028. Alongside this, AWS launched AI Spring Singapore, a flagship collaboration with government and industry aligned to National AI Strategy 2.0, aiming to train 5,000 individuals in AI annually from 2024 to 2026.
AWS is not alone. Universal AI Platform Dataiku, which has established its regional headquarters in Singapore, powers blue-chip organisations across the region – including Standard Chartered, Maybank, Auckland Transport and Security Bank in the Philippines. Beyond commercial success, Dataiku has supported Nanyang Technological University's pioneering Early Alert for Learning Intervention project, which achieved over 70 per cent accuracy in identifying students who genuinely needed help – improving retention and academic outcomes.
In other words, global AI leaders continue to bet on Singapore as a trusted digital node in a fracturing world. Now, a new frontier is coming into focus: Agentic AI.
Agentic AI as digital backbone
AI is no longer just a tool waiting for instructions – Agentic AI marks the emergence of systems that act with purpose, autonomy and impact.
Unlike traditional AI systems that wait for human prompts, Agentic AI refers to autonomous systems that make decisions, take actions and adapt in real time to achieve defined goals. Picture intelligent agents that don't just detect financial anomalies but launch mitigation steps, or healthcare systems that not only identify at-risk patients but automatically schedule interventions, adjust treatment protocols and coordinate with care teams. This shift from passive tools to active, goal-oriented systems signals a profound change in how organisations will operate with greater efficiency and effect.
Yet with autonomy comes complexity. As agents proliferate across business functions, organisations face mounting governance and security risks. Models evolve rapidly, costs are difficult to monitor, and without oversight, enterprises risk compliance gaps, technical debt and uncontrolled access to data. IT teams are increasingly caught between building their own guardrails, locking into single vendors or permitting open exploration – each with risks to scale and control.
To become a true digital backbone, Agentic AI must be deployed with governance at its core. This means centralising agent creation, ensuring rigorous oversight, and embedding tools for continuous performance and cost monitoring from the outset. Globally, pilot programmes in healthcare, finance and education already report up to 40 per cent efficiency gains and 30 to 50 per cent time savings in administrative functions.
Unsurprisingly, the Agentic AI market in Asia-Pacific is expected to grow at a staggering compound annual rate of almost 50 per cent, surpassing S$10 billion by 2030. In Singapore, demographic shifts are accelerating. In fact, one in four citizens will be 65 years of age and above by 2030, and the old-age support ratio will fall from 3.7 today to just 2.7. This means Agentic AI is not just an opportunity but a necessity. As labour-driven growth comes under increasing strain, there is a growing imperative for technologies that can enhance productivity and service quality without adding to headcount pressures.
Agentic AI isn't just the next frontier in automation – it's the operating layer for a more resilient, efficient and adaptive enterprise.
Over 60 years, Singapore has weathered crisis after crisis, emerging stronger each time through foresight, unity and skill. As the rules-based global order frays, this is not the time to retreat, but to double down on the digital foundations that have been laid. And in tech's next chapter, Agentic AI stands out as an enabler of national resilience, relevance and reach in a more uncertain world.
The writer is senior vice-president and general manager of Asia Pacific & Japan for Dataiku