Singapore is keeping pace with the rest of the world in deploying artificial intelligence (AI) but is falling well short when it comes to generating measurable returns. That gap is now translating into the lowest AI budget allocations among major markets surveyed.
Only 33% of Singapore-based respondents report quantified returns on their generative AI investments, compared to 49% globally, according to a report by Snowflake and research firm Omdia by Informa TechTarget. The study surveyed 2,050 business and technology leaders across 10 countries.
The findings point to a disconnect between activity and outcomes. While 39% of Singapore respondents say their organisations have deployed generative AI across many use cases, matching the global average, the technology is being applied less consistently across departments.
Singapore organisations lag global counterparts across every major business function tracked, including cybersecurity, where local adoption stands at 33% versus 55% globally, and IT operations, at 45% against 63% worldwide.
That narrower deployment footprint appears to be weighing on results. Fewer Singapore respondents report gains in operational efficiency (79% versus 89% globally) and cost reduction (75% compared to 82%).
A significant part of the problem may be data. Singapore has the worst data-readiness score of any country in the survey, with 73% of local respondents saying 10% or less of their unstructured data is AI-ready, compared to a global average of 40%. Without clean, accessible data, even well-resourced AI deployments struggle to generate consistent returns.
See also: Alibaba hikes AI prices as much as 34% to meet demand surge
The ROI shortfall is feeding into investment decisions. Singapore respondents estimate generative AI will account for 15% of their technology budgets over the next 12 months, the lowest figure across all 10 markets in the study and well below the 23% global average. India leads at 28%, with the UK, US, Canada and Germany all at 23%.
Part of the problem may also be strategic. While 32% of Singapore respondents identify finding specific use cases as a top challenge, only 19% of global respondents say the same. This suggests local organisations are struggling earlier in the implementation process than their peers elsewhere.
"Deploying AI and generating results are two different things. The recurring challenge is not a lack of vision, but the need for a trusted data foundation to power it," says Jenny Koh, Snowflake's country manager for Singapore.
See also: NCS, Nvidia to drive sovereign agentic AI adoption in regulated sectors
The findings arrive as Singapore's government has moved to sharpen its AI agenda. The National AI Strategy 2.0 shifted national focus from adoption to demonstrable impact, a pivot the data suggests remains a work in progress for much of the private sector.
One area where Singapore bucks the trend is agentic AI, which is the next wave of autonomous systems capable of independent decision-making and taking action without human supervision.
Some 35% of Singapore respondents say they are already using agentic solutions in production, slightly above the 32% global average. That figure sits alongside 17% who have definite plans but are unlikely to deploy within the next 12 months, versus 12% globally. This suggests a concentration of organisations either already in or preparing for a later entry, with fewer in between.
However, governance remains a concern. Six in 10 Singapore respondents acknowledge using non-approved AI tools for work, slightly above the 57% global average.
Local workers are also the most likely globally to cite approved tools lacking specific capabilities as a reason for going off-script (69% versus 60% worldwide) and 64% point to approval processes being too slow or too restrictive, compared to 53% globally. The pattern suggests that official AI programmes are not moving fast enough to meet employee demand on the ground.
Globally, the report offered a more optimistic read on AI's commercial returns. Organisations report earning roughly US$1.49 for every dollar invested, and 92% of early AI adopters say they have seen positive returns.
Still, 96% of respondents say they continue to face significant challenges in scaling their initiatives, with data quality, integration with legacy systems and employee skills cited as the most common barriers.
To stay ahead of the latest tech trends, click here for DigitalEdge Section
The workforce picture is similarly more nuanced than popular narratives suggest. Across the study, 77% of organisations report AI-driven job creation, against 46% that report role reductions.
Among organisations that have seen both, 69% say the overall workforce impact has been positive. In Singapore, 52% of respondents report AI has created jobs at their organisation, broadly consistent with the global trend. Technical functions globally are seeing the strongest net gains: 56% of IT operations teams report job growth, followed by cybersecurity at 46% and software development at 38%.
Data readiness is emerging as the primary constraint on scaling globally, with only 7% of all respondents saying more than half of their unstructured data is AI-ready. Singapore's position at the bottom of that ranking makes the challenge particularly acute locally.
