Worker vs Enterprise AI Adoption in Singapore

Singapore is routinely described as having "60%+ AI adoption," a figure that measures individual workers using AI tools, not enterprises deploying AI in their operations. Government data puts operational, firm-level AI adoption at 23.5% to 28.5%, meaning most Singapore companies have not yet formally adopted AI at all. Confusing these two numbers doesn't just misstate a statistic. It leads to the wrong assumption about who is actually buying GEO services today, and who represents unclaimed market opportunity.

By Tessar Napitupulu, Founder & CEO of PT Arfadia Digital Indonesia and Forbes Agency Council member.

Two Surveys, Two Populations, Two Very Different Numbers

The Infocomm Media Development Authority's Singapore Digital Economy Report 2025 found that 73.8% of surveyed Singapore workers use AI tools at work, most of them several times a week or daily. Separately, Microsoft's AI Economy Institute ranked Singapore second globally for generative AI diffusion, at 63.4% of the working-age population having used AI tools, trailing only the UAE at 70.1%.

These are worker-level figures. They describe individuals opening ChatGPT, Copilot or a similar tool on their own initiative, or with informal employer support, to help with a task. They say almost nothing about whether the organisations employing those workers have formally adopted AI into their operations, budgets, governance structures or customer-facing products.

That second, harder question is measured separately, and the answer is smaller. Singapore's Senior Minister of State for Digital Development and Information, Tan Kiat How, stated at ATxEnterprise 2026 that enterprise AI adoption rose to 23.5% in 2025, up from 4.3% in 2023. The Ministry of Manpower's own 2026 report put company-level AI adoption at 28.5%, meaning 71.5% of firms had yet to adopt AI in their operations at all. A third source, Morgan Stanley, reported over 70% of Singapore companies had adopted AI, a figure that looks contradictory until the methodology is examined: Morgan Stanley surveyed larger, listed firms using a broad "any AI use" definition, while the government figures measure structured, operational enterprise adoption across firms of all sizes. Different populations, different definitions, and neither number is wrong. They simply cannot be merged into one statistic without losing the distinction that makes either of them useful.

Why This Distinction Matters for a GEO Pitch

High individual AI usage is the demand-side reason GEO matters at all: it means a large share of the professionals inside a Singapore buying committee are personally comfortable asking ChatGPT or Copilot a research question, and increasingly do so before contacting a vendor's sales team. But that comfort with the tool is not the same as the organisation having a GEO strategy, an AI governance policy, or even awareness that its own visibility inside AI-generated answers can be measured and improved.

Treating 63% or 73% adoption as evidence that "everyone is already doing GEO" leads to a pitch built around urgency that doesn't match reality. Treating the 28.5% enterprise figure correctly, as evidence that 71.5% of Singapore firms have not yet formally adopted AI operationally, leads to a more accurate and, for a GEO agency, more useful framing: most of the addressable market hasn't started, but the individual habits of the people inside that market are already primed to reward the agency that gets there first.

The Global Evidence Behind "Buyers Research With AI Before Calling Sales"

Singapore-specific B2B buyer-behaviour data is thin, so it's worth being explicit that the strongest evidence for the underlying claim, that buyers now research vendors with AI tools before contacting sales, comes from global rather than Singapore-specific surveys. Gartner's survey of 645 B2B buyers, conducted August to September 2025, found 45% had used generative AI during a purchase process, primarily to gather vendor and product information, and that buyers consulted an average of seven information sources before deciding. 6sense's 2025 Buyer Experience Report put the figure considerably higher: 94% of B2B buyers reported using large language models somewhere in their buying process. The gap between 45% and 94% likely reflects different survey populations and different definitions of "used AI," a pattern that should feel familiar by now given the adoption-statistic conflation problem discussed above.

The honest version of this claim also needs its counter-trend stated, because a market this sophisticated will ask the uncomfortable follow-up question if a pitch doesn't raise it first: Gartner separately predicts that by 2030, 75% of B2B buyers will actively prefer sales experiences that prioritise human interaction over AI-driven ones, and 69% of buyers today already prefer to validate AI-generated insights with a human sales representative before acting on them. This doesn't contradict the case for GEO. It sharpens it: buyers are using AI to narrow a shortlist and form an initial impression, then still expect a human conversation to close the gap AI generated. A brand invisible to AI never makes that shortlist in the first place, regardless of how good its eventual human sales conversation would have been.

One further volatility data point is worth building into any Singapore GEO business case, because it undercuts the idea that a GEO strategy, once built, stays static: the overlap between traditional Google search results and Google's AI Overviews content dropped sharply, from roughly 76% in mid-2025 to about 38% in early 2026, according to independent tracking studies. A brand ranking well in classic Google search results is now meaningfully less likely than it was less than a year earlier to have that ranking reflected inside the AI Overview a Singapore buyer actually reads. Strong traditional SEO is necessary groundwork for GEO, but this shift is direct evidence it is no longer sufficient on its own.

Where the Real Adoption Is Concentrated

Sector-level data sharpens this further. The Ministry of Manpower's 2026 figures show adoption is highly uneven by industry: Information and Communications leads at 74.1%, followed by Professional Services at 57.5%, and Financial and Insurance Services at 56.4%. Firm size matters as much as sector; adoption reaches 76.4% among firms with more than 500 employees, compared with 23.9% among firms with fewer than 25 employees.

Among firms that have adopted AI, the pattern of use skews toward accessible, low-friction tools rather than custom infrastructure: 84% rely on off-the-shelf generative AI tools, 52% use domain-specific AI-enabled solutions, and 44% have implemented customised or proprietary tools. Larger firms use AI across a wider range of business functions (an average of five) than SMEs (an average of three), concentrated in IT, customer service, and finance and accounting.

Among AI-adopting firms surveyed by the Ministry of Manpower, more than two-thirds intend to prioritise training and upskilling their workforce over the next one to two years, and 63% expect to redesign jobs to integrate AI more deeply into daily operations. That trajectory matters for a GEO pitch specifically: a firm actively investing in AI upskilling is a firm more likely to already have, or soon acquire, the internal literacy to evaluate a citation-rate methodology on its own merits, rather than needing the concept of AI-search visibility explained from first principles.

This is a useful account-prioritisation map for any agency entering this market. It doesn't say GEO only matters to these sectors; it says these sectors are where the operational, governance-aware AI conversation is already happening, which tends to be where a GEO pitch gets a faster, better-informed hearing rather than a first-principles explainer.

Why the Enterprise Number Is Likely to Move Faster From Here

The 71.5%-not-yet-adopted figure is a snapshot, not a ceiling, and the scale of institutional investment behind Singapore's National AI Strategy 2.0 suggests the enterprise adoption curve has more acceleration ahead of it than the worker-adoption curve does, precisely because worker adoption is already close to saturating in the sectors that lead it. The Singapore government has committed more than S$1 billion to public AI research and talent development from 2025 onward. The Enterprise Compute Initiative, a joint programme between the Economic Development Board, the Digital Industry Singapore unit, Google Cloud, Microsoft and AWS, offers eligible Singapore companies up to S$500,000 to S$600,000 in cloud credits and implementation support specifically to build internal AI Centres of Excellence. Separately, AWS alone has committed to investing US$9 billion in Singapore's cloud infrastructure by 2028.

None of this guarantees any individual firm adopts AI faster. What it does establish is that the current enterprise adoption gap is not a funding or infrastructure problem for Singapore specifically, the way it might be in a less developed market. The constraints firms report, integration complexity and data security, are organisational and governance problems, which is a different, more tractable kind of barrier for a services agency to help solve than a lack of underlying infrastructure would be.

What This Means for How a GEO Engagement Should Be Framed

Two of the most-cited constraints on deeper enterprise AI adoption in Singapore are integration complexity, at 56.1% of larger firms, and data security concerns, at 55.4%. Neither of these is a GEO-specific objection, but both shape how a GEO proposal lands. A pitch that treats GEO as a narrow content deliverable, disconnected from governance, security and workflow ownership, will read as naive to a procurement team already wrestling with exactly those constraints elsewhere in its AI rollout. A pitch that explicitly addresses data governance, PDPA-readiness and workflow integration alongside the content and measurement work tends to land better, because it matches the vocabulary the buyer is already using internally.

The practical implication for market entry: lead with the 71.5%-of-firms-not-yet-adopted framing rather than the 60%+-worker-adoption framing. It's a less flattering headline, but it's the more accurate description of where the actual buying opportunity sits, and Singapore's own sophisticated enterprise buyers are more likely to trust an agency that gets this distinction right than one that repeats the flattened, merged statistic every other vendor uses.

The broader question of how AI adoption statistics translate into GEO market opportunity, across multiple Asia-Pacific markets, is covered in Cited or Silent.

Frequently Asked Questions

Is Singapore's AI adoption rate really lower than the "60%+" headline suggests?

Individual worker adoption genuinely is around 63% to 74%, depending on the survey. What's lower is operational, enterprise-level adoption, at 23.5% to 28.5%. Both figures are accurate; they simply measure different things, and only the second one describes organisational readiness to buy structured AI-related services.

Why do Morgan Stanley's figures look so different from the government's?

Morgan Stanley surveyed larger, listed companies using a broad definition of "any AI use." Singapore's government figures measure structured, operational adoption across firms of all sizes. Neither is incorrect; they answer different questions about different populations.

Which sectors should a Singapore GEO agency prioritise first?

Information & Communications (74.1% adoption), Professional Services (57.5%) and Financial & Insurance Services (56.4%) show the highest operational AI adoption, making them the sectors most likely to already have the internal vocabulary and governance awareness for a GEO conversation.

Does low enterprise adoption mean GEO isn't worth pursuing in Singapore yet?

The opposite. High individual worker comfort with AI tools means the buyer-research behaviour GEO responds to is already happening. Low enterprise adoption means most competitors haven't formalised a response yet, which is exactly the kind of gap a market-entry strategy should target.

What stops larger Singapore firms from adopting AI more deeply?

Integration complexity (56.1%) and data security concerns (55.4%) are the two most-cited constraints among larger firms, according to the Ministry of Manpower's 2026 report. Both point toward governance and security readiness as prerequisites for any AI-adjacent service pitch, not just GEO specifically.

Do B2B buyers actually still want to talk to a human, or has AI replaced that entirely?

Both are true at once. Global surveys show a large majority of B2B buyers already use AI tools somewhere in their research process, but Gartner also predicts 75% of buyers will prefer human-led sales interactions by 2030, and 69% already prefer validating AI-generated insights with a human rep today. AI shapes the shortlist; humans still tend to close the sale.

For a deeper look at how AI-adoption data should inform GEO market entry across Asia-Pacific, Cited or Silent covers this in more depth. Get the free excerpt here, or explore Arfadia's GEO & AEO service for Singapore-market engagements.