In this guide · 7 sections
Short answer
In 2025, 23.4% of Singapore SMEs had adopted AI, up from 14.5% in 2024 (IMDA SGDE Report 2026, p. 11). Close to 9 in 10 AI-adopting firms use off-the-shelf tools such as ChatGPT or Copilot, and the most cited benefit is productivity (87.6%).
The key figures from the SGDE Report 2026
The report is the fourth edition of IMDA's annual Singapore Digital Economy (SGDE) study. IMDA announced it in a press release dated 5 October 2026, and the full report PDF runs to 31 pages. Figures below refer to calendar year 2025 unless stated.
| Figure | Value | Report page |
|---|---|---|
| Digital economy value added | S$144.1 billion, 19.3% of GDP (18.8% in 2024) | p. 3 |
| SME AI adoption rate | 23.4% (14.5% in 2024) | p. 11 |
| Large local firm (non-SME) AI adoption | 70.4% (62.5% in 2024) | p. 11 |
| AI adoption, firms with 10+ staff | 34.6%, against an OECD average of 20.3% | pp. 12-13 |
| Firms planning AI in the next 12 months | 12.1% of non-adopters (5.6% in 2023) | p. 15 |
| SME digital adoption intensity | 2.5 of 6 digital areas (2.3 in 2024) | p. 11 |
The six digital areas IMDA measures are cybersecurity, cloud, e-payments, e-commerce, data analytics and AI. An average SME now uses two or three of them.
What SMEs are actually adopting: mostly off-the-shelf AI
The most useful part of the report for a small business is how firms are using AI. On page 15, IMDA reports that among AI-adopting firms:
- 89.6% use commercial off-the-shelf AI tools with no customisation. The press release gives ChatGPT, Claude, Gemini and Microsoft Copilot as examples.
- 28.3% use commercial tools with some customisation, such as configured agents or apps built on AI APIs.
- 23.0% use AI tools developed in-house or outsourced for firm-specific needs, more common in medium and large firms and the tech sector.
- 2 in 5 say they are expanding AI across the company or have fully integrated it. The rest are testing or on their first use case.
Firms can use more than one type, so these do not add up to 100%. The practical reading: most of your peers who "use AI" are paying for a chat assistant, not building models.
Benefits reported (p. 16) are mostly about efficiency: 87.6% cite productivity and process improvement and 44.0% cite cost reduction. Fewer cite revenue: 27.8% say it helped acquire new customers and 24.5% say it contributed to revenue growth.
Why many small firms still are not using AI
Page 14 lists the top barriers among firms that have not adopted AI. The leading reason has barely moved, while skills and IT barriers have eased.
| Barrier | 2023 | 2025 |
|---|---|---|
| Do not see the need due to small business | 57.4% | 54.7% |
| No budget or investment too high | 35.1% | 29.2% |
| IT infrastructure not ready | 39.5% | 26.2% |
| Lack of skills or talent | 30.8% | 24.1% |
| Poor awareness of the technology | 25.6% | 19.1% |
"Too small to need it" is sometimes right. A three-person trading company does not need an AI strategy. But it often means nobody has looked at which repetitive jobs (quotes, replies, invoices, translation) take hours each week.
Workers already use AI: the training and policy gap
The worker survey is the part most owners miss. According to the press release, 86% of surveyed workers used AI at work in 2026, up from 78% in 2025. Content creation is the most common use (76.7% of AI-using workers, p. 18), followed by process automation at around two-thirds.
Training lags behind. About 68% of workers said they need to upskill in AI, but only 36.7% had attended AI-related training in the past 12 months (p. 20). Around 35% of those without training said their employer had not nominated them.
On page 20, only around half of AI-using workers said they frequently or always fact-check AI output or hold back from sharing sensitive personal data with AI tools. For a business, that is the real risk: staff pasting customer details into a free consumer account. A short AI use policy and a business-plan account fix most of it. Our guide to a ChatGPT at work policy under the PDPA has a template.
Practical next steps by business size
The report does not prescribe actions, so the steps below are our suggestions based on its findings. Pick the row closest to your business.
| Business size | Sensible next step | Typical first use |
|---|---|---|
| 1-5 people | One business AI subscription per person, with training off and a one-page rule on customer data | Drafting replies, quotes and translations |
| 6-20 people | Written AI policy, a shared prompt library and one automated workflow | Enquiry follow-up, invoice reminders, document extraction |
| 21-50 people | Named owner for AI tools, access review, and a pilot of a customised tool on your own documents | Internal knowledge assistant, website assistant |
List the five most repetitive weekly tasks
Ask staff where they already use AI. The survey suggests many do, with or without permission.
Choose a business plan, not free accounts
Business plans of ChatGPT, Claude, Gemini and Copilot let you control data use and remove access when someone leaves.
Automate one workflow end to end
Pick a task with clear rules, such as replying to web enquiries, and measure time saved over a month.
Check funding before you buy
IMDA announced free AI trials and a curated tool list in October 2026; see our summary of IMDA's AI support for SMEs. For grant-funded projects, read our EDGE Grant guide.
Do it yourself or get help?
Most of the steps above are owner decisions, not IT projects: choose a plan, write the rules, show staff what is allowed. Where outside help pays off is the first automation or customer-facing assistant, because that is where data handling and error checks matter.
We set up website assistants that answer from your own content as an AI Website Assistant package (S$790), and build single workflows such as enquiry follow-up or invoice reminders as Workflow Automation (S$390 per workflow).
Frequently asked questions
IMDA's Singapore Digital Economy Report 2026 says 23.4% of SMEs had adopted AI in 2025, up from 14.5% in 2024. Among large local enterprises the rate was 70.4%. Across all enterprises it was 23.5%, and 34.6% among firms with ten or more employees.
IMDA estimates Singapore's digital economy at S$144.1 billion in value added in 2025, or 19.3% of GDP, up from S$136.5 billion and 18.8% in 2024. About two-thirds comes from digitalisation in sectors outside information and communications, such as finance and wholesale trade.
Most AI-adopting firms use off-the-shelf tools. The report says 89.6% use commercial tools with no customisation, with ChatGPT, Claude, Gemini and Microsoft Copilot named as examples. Some 28.3% customise commercial tools and 23.0% use tools built in-house or outsourced.
IMDA publishes the report as a PDF linked from its 5 October 2026 press release titled Singapore's Digital Economy Continues to Grow Reaching 19.3%. The report has sections on digital economy size, tech manpower, enterprise digitalisation, AI adoption and AI usage among workers.
It is the most common reason non-adopters give, cited by 54.7% in 2025. Small firms rarely need custom AI, but most have repetitive writing, reply or data-entry tasks where a business AI subscription or one simple automation saves time. Start with one task and measure the result before spending more.