Tech Week Singapore: From Policy to Pilots, Asia’s AI Momentum Takes Shape
What we heard on the mainstage and why it matters for APAC
Asia Tech Lens was on the ground at Marina Bay Sands for Day 1 of Tech Week Singapore - an umbrella of six mega-shows spanning AI, cloud, data centers, cybersecurity and e-commerce.
The scale hits you the moment you enter: robots greeting delegates, power-hungry racks humming in the data-center hall, and executives talking regulation, reliability and ROI over back-to-back demos and panels.
If 2023 was the year of hype, 2025 feels like the year of deployment. From Singapore’s national AI playbook to enterprises building “agent teams,” the tone has shifted from “what’s coming” to “what’s working” and “what needs to change”.
Here’s what stood out and what’s next for the region.
1. Singapore’s AI Playbook: Rules, Adoption, and Talent
The opening keynote by Senior Minister of State Tan Kiat How at the Ministry of Digital Development and Information, outlined Singapore’s three-pillar AI strategy:
Regulations and guidelines about developing AI and the use of AI.
Encouraging adoption of technology, especially AI, across all sectors.
Talent and manpower development.
He pointed to the Monetary Authority of Singapore (MAS) FEAT principles and the Veritas toolkit as examples of Singapore’s “regulation with humility”: co-developing guidelines with industry before stepping in with legislation, and only when clear risks emerge.
The numbers back the ambition:
The digital economy now contributes 18.6% of GDP.
AI adoption among SMEs has jumped from 4% to 15% in 2024.
Among larger firms, adoption has risen from 40% to 60%.
The AI workforce grew 25% last year, targeting 15,000 professionals.
ATL’s Take:
Singapore’s AI strategy follows its classic playbook, build talent pipelines early, codify clear rules, and let industry co-design the framework. The approach is less about chasing moonshots and more about making AI usable, safe, and scalable across every sector.
2. From Single Agents to Agent Team
The next session, AI & the C-Suite: Leading the Charge into the Future of Innovation, was led by Sachin Chitturu, Partner and Southeast Asia Leader at QuantumBlack, McKinsey’s AI arm known for taking models into production.
What we heard
From solo bots to agent teams: Companies are moving from pilots with single agents to teams of agents coordinating on research, customer service, software development, and fraud detection.
Two-layer roadmap: “Passive AI” (analytics and optimization) still dominates, but “Active AI” (reasoning and adaptive systems) is emerging. Most firms lack the data maturity for either layer to thrive.
Common blockers: 60% of AI projects stall on data integrity. Less than 10% of data feeds operate in real time, and under 30% of decisions have outcome labels - meaning the systems don’t know if they’re right.
Infrastructure tax: Scaling often triggers 10-20× jumps in compute demand.
New oversight: Human-AI “assembly lines” are forming - teams where agents handle repetitive loops while humans govern feedback and learning cycles.
ATL’s Take:
Chitturu’s framing is spot on, the next wave of enterprise value will come from collaborating agents, not isolated copilots. But most firms are still at base camp: data pipelines are dirty, outcome feedback loops thin, and infra costs steep. The leap from pilot to production is less about models, more about organizational plumbing.
3. APAC’s GenAI Charge
OpenAI’s Andy Brown and Accenture’s Dr. Ramine Tinati discussed how generative AI has moved from pilots to day-to-day work in Asia.
What we heard
OpenAI now reaches 800 million weekly users and 5 million businesses, with Asia leading global usage growth.
Singapore is OpenAI’s regional hub, a strategic anchor for both enterprise partnerships and talent sourcing.
Case studies:
Air New Zealand empowered all staff to automate tasks, cutting process times from days to seconds.
CNA Newsroom in Singapore overcame editorial hesitation by allowing guided experimentation, improving both efficiency and story quality.
The central message: AI is not a replacement for people, it’s a multiplier.
ATL’s Take:
Brown’s examples show that when experimentation is sanctioned, productivity soars. The challenge now is trust governance: who sets the sandbox rules so creativity thrives without crossing policy or brand lines?
4. Beyond Pilots: Turning AI into Trusted Systems
A heavyweight panel tackled the question: How can APAC turn AI pilots into durable, trusted systems?
Moderated by Prof. Keith Carter, KDA Capabilities, the discussion brought together:
Dr. David R. Hardoon, Global Head, AI Enablement - Standard Chartered.
Dr. Ott Velsberg, Estonia’s Chief Government Data Officer – the state platform builder.
Dr. Aik Beng Ng, NVIDIA AI Technology Center – the compute ecosystem voice.
What we heard:
Jobs evolving, not vanishing: Experience in Estonia shows most roles transform rather than disappear.
Trust as the foundation: Education, inclusion, privacy, and transparency remain the four prerequisites.
Sovereign and local AI: Smaller nations can thrive by designing AI in their own languages and contexts - sovereignty as adaptation, not isolation.
Public–private partnerships: Practical enablement beats premature regulation.
Security realism: Cyber threats move faster than policy - iteration matters more than perfection.
ATL’s Take:
If AI is to deliver real value, APAC needs a common grammar, between regulators, enterprises, and citizens. The real opportunity isn’t disruption; it’s connection. The real risk isn’t change; it’s fragmentation.
5. The WTO’s Lens: AI as a Trade Engine
Capping off Day 1, Emmanuelle Ganne, Chief of Digital Trade and Frontier Technologies at the World Trade Organization, zoomed out to the global stage. Her message: technology’s true promise is inclusivity - using AI to make trade more efficient and accessible, not just faster.
What we heard:
Big upside: WTO simulations show AI could lift cross-border trade by ~40% by 2040 and boost global GDP by 12–30%.
SMEs are bullish: ~70% expect AI to cut trade costs; 90% already see benefits. Small firms are more optimistic than large ones.
Three major risks:
Persistent digital divides, with US $418 billion needed to connect the unconnected.
Skills gaps: Only one-third of low-income countries have AI education strategies.
Fragmentation: a two-bloc digital split could shrink world GDP by 7%.
Standards, not silos: WTO backs interoperable data standards and regional resource pooling (e.g. shared data centers) over national isolationism.
Current pulse: Trade is projected to grow 2.4% this year, boosted by AI-enabling goods up 20% YoY.
ATL’s Take:
AI could be the next great trade equalizer - but only if the pipes connect. For APAC, the homework is clear: fix connectivity, close skills gaps, and align data rules. The danger isn’t lagging in AI models - it’s building them on disconnected islands.
The Signal
Tech Week Singapore showcased a region moving from experimentation to execution.
Governments are codifying trust frameworks.
Enterprises are scaling agent ecosystems.
Global institutions are linking AI to trade and inclusion.
Asia’s playbook for 2025 is emerging:
Pragmatic, iterative, and anchored in people and not just platforms.
For More Info on Asia Tech Lens