Singapore SMEs have adopted AI. New AWS research maps the maturity gap that follows

Across Singapore’s financial services, healthcare, and manufacturing sectors, SMEs are no longer asking whether AI matters. In many cases, they are already using it. The more important question now is what comes after adoption — and whether organizations have the structures in place to turn early momentum into durable, organization-wide value.

That is the picture emerging from the new AWS-commissioned “Unlocking Singapore’s AI Potential” research. Conducted by research advisory firm, Strand Partners, the report examined AI adoption among Singapore SMEs in three of the country’s priority sectors. While AI use is already widespread, far fewer organizations have moved into more advanced, integrated use across core functions. In financial services, 75% of SMEs surveyed say they are using AI, but only 29% of those adopters say they have progressed to advanced use, where they are combining multiple AI models or tools, or creating their own AI systems. In manufacturing, 57% say they are using AI, but only 23% have reached that stage. In healthcare, these figures are 61% and 16% respectively.

"Singapore's SMEs have moved decisively on AI. The next step is making that investment sustainable by integrating AI across functions so it shifts from a point solution to a core part of how the businesses run," said Priscilla Chong, Managing Director, Singapore, AWS. 

The challenge starts after the first use case

The findings suggest the challenge has moved from getting AI started to getting it embedded. And where it plateaus depends on the sector. In financial services, 38% of SMEs say internal approval and sign-off is the longest stage in AI deployment. In healthcare, 30% say the same. In manufacturing, the longest stage is not approval but systems integration with existing workflows, cited by 37% of businesses.

“These are different pressure points, but they signal progress. Businesses have moved past the starting line. Now the work is about shaping the culture and processes to match the ambition,” said Ms Chong.

AI stewardship exists. It just hasn't been democratized yet 

Across all three sectors, at least half of SMEs say humans retain final stewardship or share final decision-making with AI systems, but the formal mechanisms are still taking shape. Almost 30% of AI-adopting SMEs across the surveyed sectors say they have a clearly defined person responsible for overseeing AI accuracy. In many cases, accountability is informal — shared across teams, assigned to the team that built the system, or determined case-by-case.

Stewardship is also about continuity. Six in ten SMEs across the three sectors say they would face significant or moderate disruption if the primary person responsible for AI left the organization, while around a tenth say their AI initiatives would likely stop altogether. This suggests that, in many businesses, responsibility for AI still rests heavily with a small number of individuals rather than being fully embedded in wider structures and teams.

That does not indicate a lack of intent, but is a signal of SMEs in transition. The next step is making those mechanisms clearer, more consistent, and easier to apply as AI becomes more embedded in everyday business decisions.

Working alongside SMEs in Singapore, Ms Chong sees the pattern up close. “The companies getting this right tend to share a common approach: a dual strategy. One environment that's safe and open for experimentation, where employees across the business can build familiarity and confidence with AI tools. And another with strict guardrails for production, where proven use cases run at scale against real business problems. When you compound those two approaches: a workforce gaining skill alongside a pipeline of high-value use cases, stewardship stops resting on one person's shoulders and starts becoming a shared capability. The organizations pulling ahead aren't waiting for a perfect framework before they start. They're building the muscle and the guardrails at the same time.”

AI has moved beyond the tech team 

While IT and technology functions remain the most common lead for AI initiatives, a significant share of businesses say those initiatives are also being driven by non-technical units such as marketing, HR, and operations. Across the three sectors, under 40% say they do not have a formal process for escalating AI outputs that employees are unsure about. That means the ability to question an output, pause, and seek guidance may still depend too heavily on individual confidence rather than a clear channel for raising concerns.

“The opportunity is clear. Businesses that build clear, simple escalation routes, something as straightforward as a defined step for when a team member questions an AI-generated forecast before it reaches a client, will move faster and get more from their AI investments,” added Ms Chong.

One journey, different sector needs

In financial services and healthcare, the longest stage of deployment is approval and sign-off. In manufacturing, it's systems integration with existing infrastructure. Yet regardless of where the friction sits, the downstream challenge is the same: around two-thirds of businesses that found industry-specific guidance say they still had to adapt it significantly before it became fit-for-purpose. Only a minority (20% in healthcare, 17% in financial services, and 13% in manufacturing) say the guidance they found was directly applicable to their organization.

That matters because it suggests the issue is not a lack of industry direction, but depth.

“A financial services SME navigating internal approval needs a different playbook from a manufacturer working through systems integration, but both need guidance from technology partners and industry peers that reflects how their business actually runs, not just how AI works in theory. The technology is ready. The opportunity now is building the organizational muscle, the processes, the confidence, the clarity, so more teams can put it to work. That's what turns AI exploration into operation,” said Ms Chong.

A note on methodology and disclaimers:
This research was commissioned by Amazon Web Services and conducted by Strand Partners, an independent research consultancy. The fieldwork follows the guidance set forth by the UK Market Research Society (MRS) and ESOMAR. A total of 1,500 unique businesses in Singapore were surveyed. The sample was structured in two parts: a nationally representative sample of 750 businesses, weighted by business size and sector to reflect Singapore's overall business landscape; and targeted sector deep dives in financial services (n=300), manufacturing (n=307), and healthcare (n=288) to enable more detailed analysis of SMEs in Singapore's three National AI Missions sectors. The nationally representative sample already included some financial services, manufacturing, and healthcare firms, which were counted in both the national sample and their respective sector deep dive. The remaining respondents in each sector were recruited specifically for the deep dive. For the purposes of this study, SMEs are defined by Singstat as companies with a Group Annual Sales Turnover of not more than S$100 million or a Group Employment Size of not more than 200 workers. Business leaders are defined as founders, CEOs, members of the C-suite, or senior decision-makers with responsibility for technology or digital strategy within their organization. The findings represent the views of the businesses surveyed and should not be taken as an assessment of Singapore's national AI strategy or government frameworks. SMEs were not asked to evaluate or comment on existing government policies or programs. All sector-level findings refer to the surveyed sample within each sector and are not intended as a characterization of the sector as a whole.