The Two-Speed Economy Nobody Wants to Talk About
Here's an awkward truth about Singapore's AI revolution: most employees are already using it, but most companies haven't figured out how to make it stick. A new McKinsey-EDB report found that 56% of Singapore companies are scaling AI, well above the Southeast Asian average of 38%. Impressive, until you look at who's actually doing the scaling.
Large enterprises are racing ahead. Small and medium enterprises, which make up 99% of Singapore's businesses and employ two-thirds of its workforce, are largely watching from the sidelines. As we've seen in broader regional adoption patterns, the gap isn't closing. It's widening.
By The Numbers
- 56% of Singapore companies are scaling AI, versus 38% regional average (McKinsey-EDB, February 2026)
- 63% of Singapore businesses have operationalised or partly implemented AI in 2026, up from 45% in 2025 (Gallagher Survey)
- Over 50% of Singapore businesses cite skills gaps as the primary obstacle to AI adoption
- 46% of Southeast Asian firms have scaled AI beyond pilot phases, surpassing the 35% global average
- S$37 billion committed under RIE2030 for research and capability development
Employees Aren't Waiting for Permission
The most telling data point isn't about companies at all. It's about individuals. Across Singapore's workforce, employees are quietly adopting AI tools on their own. They use ChatGPT to draft emails, Gemini to summarise meeting notes, and various AI coding assistants to speed through routine tasks.
None of this shows up in enterprise adoption statistics because it's happening outside company systems. PwC Singapore has flagged this as a structural problem. When employees use personal AI tools informally, the productivity gains disappear when they leave. The knowledge stays with the individual, not the organisation.
"Without diffusion, a two-tier economy potentially emerges with multinational corporations thriving while smaller players stall." - Sujay Shetty, Partner, PwC Singapore
This isn't hypothetical. It's already happening. Walk into any large Singapore bank or logistics firm and you'll find dedicated AI teams, custom models, and integrated workflows. Walk into a neighbourhood accounting practice or a mid-sized manufacturer and the most advanced AI tool is likely someone's personal ChatGPT subscription.
Budget 2026: The Government's Response
Singapore's Budget 2026, delivered in February, acknowledged the gap directly. The government is expanding several programmes aimed at dragging SMEs into the AI era, whether they're ready or not.
| Initiative | What It Does | Who Benefits |
|---|---|---|
| Enterprise Innovation Scheme (EIS) | 400% tax deduction on qualifying AI expenditure, capped at S$50,000 | All businesses |
| Productivity Solutions Grant | Expanded to cover wider range of AI-enabled solutions | SMEs |
| Champions of AI programme | Supports firms committing to full AI transformation | Mid-size and scaling firms |
| SkillsFuture AI courses | Six months free access to ChatGPT and Gemini for enrolled workers | Individual workers |
| Market Readiness Assistance | Double tax deduction for internationalisation, ceiling raised to S$400,000 | Exporting SMEs |
The SkillsFuture enhancement is particularly notable. By giving workers free access to premium AI tools through training programmes, the government is trying to formalise what employees are already doing informally. The logic: if individuals are going to use AI anyway, at least make sure they know how to use it well and within organisational frameworks.
Why SMEs Are Stuck
The barriers are well documented by now. The Singapore Business Federation's National Business Survey found that SMEs consistently cite three problems: no in-house expertise, no clarity on where to start, and no way to prove return on investment before committing resources. This mirrors challenges we've identified in broader transformation failures across the region.
But there's a deeper issue. Most AI tools available today were designed for large organisations with dedicated IT teams, clean data pipelines, and the budget to experiment. An SME with 15 employees and a spreadsheet-based workflow can't simply plug in an enterprise AI platform and expect results.
"End-to-end AI adoption requires data organisation, process redesign, and worker retraining. Most SMEs are still at step zero." - Gan Kim Yong, Deputy Prime Minister, Singapore (Budget 2026 address)
- Cost of entry remains high, with most enterprise AI platforms priced for large organisations rather than firms with fewer than 50 employees.
- Data readiness is the quiet bottleneck: SMEs often lack the structured, clean datasets that AI tools require to deliver results.
- Talent competition is lopsided, with large firms and government agencies absorbing most of Singapore's small but growing AI workforce.
- Risk aversion runs deep among SME owners who have survived multiple economic cycles by avoiding unproven investments.
- Integration complexity means adding AI to existing workflows often requires rebuilding processes from scratch.
The Regional Context
Singapore isn't alone in facing this gap, but it's more exposed because of how much the government has staked on AI as a national strategy. Under the National AI Strategy 2.0, Singapore has committed over S$1 billion to AI research and development through 2030. The country has launched dedicated AI missions in areas like healthcare, education, and finance.
Across the border, the story is different but equally instructive. Indonesia, with 51% of firms reportedly adopting AI, is seeing broader but shallower adoption. Vietnam is building AI capacity from a lower base but with a younger, more digitally native workforce. Malaysia is attracting data centre investment that could underpin future AI infrastructure.
A February 2026 DBS survey of 730 Singapore companies found that 82% plan to internationalise this year. For those companies, AI isn't optional. It's the difference between competing in regional markets and being left behind by more agile rivals in neighbouring countries. As our coverage of Asia-Pacific enterprise AI investment shows, the stakes are only getting higher.
The Hidden Cost of Falling Behind
What happens when half the economy races ahead while the other half watches? The answer is already visible in Singapore's labour market. Large firms are poaching AI-skilled workers from SMEs by offering significantly higher salaries. SMEs that might have competed for talent five years ago now find themselves priced out entirely.
The productivity gap compounds over time. A 2025 study by the Singapore Institute of Management found that companies with integrated AI workflows were 23% more efficient at processing customer requests than those relying on manual processes. For SMEs competing against AI-enabled rivals, this isn't just about efficiency. It's about survival.
"The AI divide isn't just about technology. It's about whether Singapore's SMEs can remain competitive in an increasingly automated regional economy." - Dr Sarah Lim, Director, AI Singapore (February 2026 forum)
This challenge extends beyond Singapore's borders. As we've documented in our analysis of Southeast Asia's AI ambitions, the region faces similar structural divides between AI leaders and laggards.
FAQ
How many Singapore SMEs have adopted AI beyond basic tools?
Precise SME-specific figures are difficult to pin down. A 2023 government study found only 4.2% of SMEs had adopted AI, compared to 44% of large enterprises. By 2026, overall scaling rates have improved, but the gap between large firms and SMEs remains significant.
What AI support does Singapore Budget 2026 offer for small businesses?
Budget 2026 expanded the Productivity Solutions Grant for AI-enabled tools, introduced 400% tax deductions on AI spending through the Enterprise Innovation Scheme, and launched six months of free AI tool access through SkillsFuture training programmes.
Why are Singapore employees adopting AI faster than their companies?
Consumer AI tools like ChatGPT and Gemini are free or cheap and require no IT approval. Employees use them informally for personal productivity, but these gains aren't captured at an organisational level because the tools sit outside company systems.
Which Singapore SMEs are most at risk from the AI gap?
Professional services firms, manufacturers with complex supply chains, and businesses competing in export markets face the highest risk. These sectors require AI integration for efficiency gains that determine competitive positioning in regional markets.
How does Singapore's SME AI adoption compare regionally?
Singapore's overall AI scaling rate of 56% leads Southeast Asia, but this figure masks the SME reality. Countries like Vietnam and Indonesia may have lower headline figures but more distributed adoption across business sizes, creating different competitive dynamics.
The question isn't whether Singapore's SMEs will eventually adopt AI. Market forces will ensure they do. The question is whether they'll do it fast enough to remain competitive, or whether they'll be relegated to serving the domestic market while AI-enabled rivals capture regional opportunities. What's your experience with AI adoption in Singapore's SME sector? Drop your take in the comments below.
No comments yet. Be the first to share your thoughts!
Leave a Comment