From 4.2% to 14.5% in Twelve Months
In 2023, just 4.2% of Singapore's small and medium enterprises had adopted AI. By the end of 2024, that figure had reached 14.5% — a tripling in a single year (Pertama Partners SEA SMB AI Adoption Index 2026). For a metric that typically moves in single-digit increments, this acceleration is remarkable. It did not happen by accident.
Singapore now leads the Pertama Partners SEA SMB AI Adoption Index at 52 out of 100, a score that is 68% higher than the regional average of 31 (Pertama Partners SEA SMB AI Adoption Index 2026). While Singapore benefits from structural advantages — a small, highly connected economy with world-class digital infrastructure — the specific policy choices, investment decisions, and ecosystem design behind this surge offer a replicable playbook that SMB leaders and policymakers across Southeast Asia should study carefully.
This is not a story about Singapore being exceptional. It is a story about what happens when a government, an ecosystem, and a business community align around the same objective.
The Three Pillars of Singapore's Acceleration
Singapore's AI adoption surge did not come from a single program or a single investment. It was the result of three mutually reinforcing pillars that addressed the barriers preventing SMEs from moving past experimentation.
Pillar 1: National AI Strategy 2.0 — Setting the Direction
Singapore's National AI Strategy 2.0 (NAIS 2.0), launched in late 2023, provided the strategic umbrella for everything that followed. Unlike many national AI strategies that focus primarily on research and large enterprise capabilities, NAIS 2.0 explicitly included SME adoption as a priority. The strategy recognized that Singapore's digital economy — now representing 18.6% of GDP, up from 14.9% in 2019 (Pertama Partners SEA SMB AI Adoption Index 2026) — would not reach its potential if AI remained concentrated among large firms and government agencies.
NAIS 2.0 created two critical conditions for SME adoption. First, it provided regulatory clarity. SMEs operating in a market where the government has clearly signaled support for AI adoption face less uncertainty about whether their investments will align with future regulations. Second, it created ecosystem coordination. By aligning IMDA (Infocomm Media Development Authority), Enterprise Singapore, SkillsFuture Singapore, and industry bodies around a common AI adoption agenda, the strategy reduced the fragmentation that often plagues government technology initiatives.
Pillar 2: SG$150 Million Enterprise Compute Initiative — Removing the Infrastructure Barrier
One of the most tangible interventions was the SG$150 million Enterprise Compute Initiative (ECI), announced in Budget 2025. The ECI addresses a barrier that is often invisible in policy discussions but decisive in practice: access to compute.
AI workloads — training models, running inference, processing large datasets — require computing resources that most SMEs cannot afford to acquire independently. Cloud compute costs, while declining, remain a meaningful expense for businesses with limited IT budgets. The ECI provides SMEs with subsidized access to compute resources, effectively removing the infrastructure cost barrier for AI experimentation and deployment.
This is significant because compute access represents a structural barrier rather than a knowledge or willingness barrier. An SME owner who understands AI, has identified a use case, and has trained their staff can still be blocked by the simple inability to afford the computing power to run the solution. The ECI closes that gap.
Pillar 3: Practical Adoption Tools — Making AI Accessible
The third pillar is a suite of practical tools and programs that reduced the friction between "wanting to use AI" and "actually using AI."
The GenAI Navigator for SMEs. Launched by IMDA, the GenAI Navigator recommends pre-approved generative AI solutions matched to specific business needs, with accompanying grant support. For an SME owner who knows they have a customer service problem but does not know which of the hundreds of available AI tools to evaluate, this is the difference between action and paralysis.
The Productivity Solutions Grant (PSG). SMEs adopting AI solutions under the PSG framework reported average cost savings of 52% in 2024 (Pertama Partners SEA SMB AI Adoption Index 2026). The PSG does not just subsidize the purchase of AI tools — it creates a structured evaluation and implementation pathway that increases the likelihood of successful deployment.
SkillsFuture Enterprise Credit. Enhanced with an additional SG$10,000 per company from 2026, this credit directly addresses the workforce training component of AI adoption. The credit can be used for AI-specific upskilling programs, ensuring that SMEs can invest in their people alongside their technology.
IMDA's Digital Leaders Programme. This program targets 2,000 local digitally mature enterprises for AI adoption support over three years, creating a cohort of SMEs that can serve as models and mentors for broader adoption.
The Results: Beyond the Headline Number
The tripling of SME AI adoption from 4.2% to 14.5% is the headline, but the supporting metrics tell a richer story about what that adoption looks like in practice.
Worker adoption is strong. Among Singaporean workers surveyed by IMDA, three in four regularly use AI tools, and 85% report that AI makes them more efficient (Pertama Partners SEA SMB AI Adoption Index 2026). This is not top-down adoption imposed by management — it is bottom-up adoption driven by workers who find the tools genuinely useful.
Cost savings are measurable. The 52% average cost savings reported by PSG participants provides concrete evidence that AI adoption delivers financial returns for SMEs, not just for large enterprises. For a market where 40% of firms cite cost constraints as a barrier to AI adoption (Pertama Partners SEA SMB AI Adoption Index 2026), having documented proof that adoption reduces costs rather than increasing them changes the calculus.
Enterprise-wide adoption is progressing. Among larger enterprises in Singapore, 62.5% had adopted AI by 2024. This creates downstream demand for AI-capable SME vendors, technology partners, and service providers — expanding the AI ecosystem beyond direct SME adoption.
What Singapore Still Gets Wrong
Singapore's acceleration should not be mistaken for a completed transformation. The Pertama Partners SEA SMB AI Adoption Index assigns Singapore a score of 52/100 — impressive relative to the region, but indicating that nearly half the journey remains.
The Implementation-Integration Gap
Singapore's dimension scores reveal where the challenge lies. The market scores 78 on Awareness and 65 on Experimentation, but drops to 48 on Implementation, 38 on Integration, and just 32 on Optimization (Pertama Partners SEA SMB AI Adoption Index 2026). The gap between "trying AI" and "embedding AI into core business operations" remains substantial even in the region's most advanced market.
The Readiness Paradox
Only 13% of organizations in Singapore feel fully prepared for AI (EY Singapore, 2025), despite the country having the highest adoption rates and most comprehensive support programs in Southeast Asia. This suggests that access to tools and training is necessary but not sufficient — organizational readiness requires deeper change management, process redesign, and leadership commitment.
Persistent Barriers
The Deloitte 2026 survey found that among Singapore-based respondents, regulations and compliance (27%), AI skills and knowledge gaps (24%), and high implementation costs (15%) remain the top challenges. The talent shortage — cited by 55% of firms regionally (Pertama Partners SEA SMB AI Adoption Index 2026) — affects Singapore as well, particularly given intense competition for AI talent from financial services firms and global technology companies based in the city-state.
How Other Markets Compare — and What They Can Learn
The value of Singapore's experience lies not in its specific programs, which are funded by a small, wealthy city-state's budget, but in the principles those programs embody. Let us examine how other Southeast Asian markets are approaching the same challenge, and where Singapore's experience offers transferable lessons.
Malaysia: The Workforce-First Model (Score: 33/100)
Malaysia has charted a distinctive approach through its Human Resource Development Corporation (HRD Corp) training ecosystem. The MDEC AI Skills Training programme is fully HRD Corp-claimable, covering AI training with no upfront fees. The National Training Week 2025/2026 targeted one million Malaysians across 70,000+ programmes (Pertama Partners SEA SMB AI Adoption Index 2026).
What Malaysia does well: Broad workforce exposure to AI tools and concepts, removing the skills barrier at the awareness and experimentation levels. Malaysia's Culture dimension score of 42 reflects the depth of its training infrastructure.
Where Singapore's experience applies: Malaysia's challenge is converting training into operational deployment. The gap between "we have trained our people" and "we have deployed AI in our operations" mirrors Singapore's earlier stage. Malaysia could benefit from a PSG-equivalent program that subsidizes implementation alongside training, and a GenAI Navigator-style tool that connects trained workers to specific, pre-vetted solutions.
Vietnam: The Velocity Play (Score: 34/100)
Vietnam is Southeast Asia's fastest mover, with a 39% year-on-year AI adoption growth rate and 47,000 new enterprises implementing AI in 2024 alone (Pertama Partners SEA SMB AI Adoption Index 2026). Vietnam's new AI law, effective March 2026, provides regulatory clarity that could further accelerate adoption.
What Vietnam does well: Speed and regulatory decisiveness. Vietnam's Experimentation score of 48 is the highest outside Singapore, reflecting genuine momentum in enterprise AI adoption.
Where Singapore's experience applies: Vietnam's adoption is broad but shallow. The country's Implementation score of 30 and Integration score of 18 suggest that many of those 47,000 new AI-adopting enterprises are in early stages. Singapore's emphasis on structured implementation support — grants tied to measurable outcomes, curated solution marketplaces, compute subsidies for scaling beyond pilots — could help Vietnam convert velocity into depth.
Indonesia: The Scale Opportunity (Score: 27/100)
Indonesia is home to 65+ million MSMEs but only 26% of organizations have implemented AI tools (Pertama Partners SEA SMB AI Adoption Index 2026). The country's AI application revenue growth of 127% from H1 2024 to H1 2025 signals strong momentum, but the base remains low.
What Indonesia does well: Sheer market size creates self-reinforcing adoption dynamics. Indonesia's "Making Indonesia 4.0" strategy provides a policy framework, and the digital economy is projected to exceed USD 130 billion by 2026.
Where Singapore's experience applies: Indonesia's most critical lesson from Singapore is the compute infrastructure investment. With 84% of Indonesian respondents citing infrastructure challenges as a top barrier, a compute initiative scaled to Indonesia's market size could be transformative. Additionally, Singapore's approach of curating pre-approved AI solutions through the GenAI Navigator is directly relevant for Indonesian MSMEs that lack the resources to evaluate a fragmented vendor landscape.
The Replicable Playbook: Five Principles
Regardless of market size, GDP per capita, or government budget, five principles from Singapore's experience apply broadly.
Principle 1: Address Infrastructure, Not Just Awareness
Most government AI programs focus on awareness and training. Singapore's insight was that infrastructure — particularly compute access — is a binding constraint for SMEs. Any market serious about SME AI adoption must invest in making AI infrastructure accessible, whether through compute subsidies, shared computing facilities, or partnerships with cloud providers.
Principle 2: Curate the Solution Landscape
The AI vendor market is overwhelming. Hundreds of tools claim to solve every business problem. SMEs lack the time and expertise to evaluate these options. Singapore's GenAI Navigator demonstrates the value of a curated, government-vetted marketplace that matches SMEs with specific, pre-approved solutions. This model can be replicated at low cost in any market.
Principle 3: Subsidize Implementation, Not Just Adoption
The PSG's 52% cost savings figure is powerful because it ties subsidies to outcomes. Rather than simply subsidizing the purchase of AI tools (which enables experimentation but not implementation), tie financial support to measurable deployment metrics: processes automated, efficiency gains documented, staff trained and using tools regularly.
Principle 4: Build Proof Points That SMEs Trust
SME owners are pragmatic. They are more influenced by evidence from businesses like theirs than by government statistics or vendor promises. Singapore's approach of documenting PSG outcomes — the 52% cost savings figure, the 85% worker efficiency improvement — creates proof points that other SMEs can evaluate.
Principle 5: Invest in Workforce Confidence, Not Just Workforce Skill
Singapore's finding that 85% of AI-using workers report efficiency gains matters because it demonstrates that AI adoption improves working life rather than threatening it. Across Southeast Asia, cultural resistance and fear of displacement remain significant barriers. Programs that help workers experience AI as a collaborative tool — rather than a replacement threat — accelerate organizational adoption.
What This Means for SMB Leaders Outside Singapore
If you are leading an SMB in Malaysia, Vietnam, Indonesia, Thailand, the Philippines, or Hong Kong, the temptation is to dismiss Singapore's experience as the product of a small, wealthy city-state with resources your market cannot match. That would be a mistake.
The underlying lesson is not about budget. It is about sequence. Singapore's success came from addressing barriers in the right order: infrastructure and compute access first, then curated solutions, then subsidized implementation, then workforce development. Many markets are doing these steps out of order or skipping steps entirely — training workers without providing compute access, subsidizing tool purchases without implementation support, or promoting AI awareness without curating the solution landscape.
For individual SMBs, the most actionable takeaway is this: do not wait for your government to replicate Singapore's ecosystem. The tools available today — cloud-based AI services, SaaS platforms with embedded AI, low-code/no-code AI builders — allow any SMB with an internet connection to begin the pilot-to-platform journey. Government programs accelerate the process, but they are not prerequisites for it.
The region's digital economy now exceeds USD 300 billion in gross merchandise value (Pertama Partners SEA SMB AI Adoption Index 2026), and Asia-Pacific AI spending is projected to reach USD 175 billion by 2028 at a 33.6% compound annual growth rate. The opportunity is expanding faster than most SMBs realize.
For the complete analysis including country-by-country scores and methodology, read the full SEA SMB AI Adoption Index 2026.
Ready to move your organization beyond AI experimentation? Book a consultation with Pertama Partners.
Frequently Asked Questions
The tripling was driven by expanded SkillsFuture funding, easier access to AI tools through the Productivity Solutions Grant (PSG), growing awareness from successful case studies, and the competitive pressure of early adopters demonstrating measurable gains in productivity and revenue.
Key lessons include: government subsidies dramatically accelerate adoption when they are easy to access, showcasing local success stories is more effective than global case studies, and bundling AI training with implementation support helps SMBs bridge the knowledge-to-action gap.
Financial services, professional services, and logistics lead in SME AI adoption. Retail and F&B are growing fastest due to customer-facing AI applications. Manufacturing lags slightly due to legacy infrastructure but is accelerating through PSG-supported automation projects.