Original research · The full report
How Asia-Pacific founders actually use AI in 2026
The complete findings from founders and business leaders across 16 markets: how deep AI runs, what they pay, which tools they reach for, and what they get back.
By Dhawal Shah · dhawalshah.net
A note to respondents
Thank you for the data
If you took this survey, this report is the promise kept. You gave honest answers about how your company really uses AI, what it costs you, and what is genuinely hard about it. In return, here is the full picture you helped build, with nothing held back behind a paywall.
What follows is the complete read: who responded, the headline findings, every chart, the patterns that show up when you cut the data by company size and market, and the obstacles you named in your own words. Where a segment is too small to be precise, it is labelled as directional rather than dressed up as a headline number.
Want to slice it yourself, or download a copy to keep? Both are at the end of this report.
A living report · Updated 1 September 2026
This is not a one-off. The data is refreshed regularly as new companies take part, and in November every respondent is re-surveyed to capture what has changed. If you run an Asia-Pacific startup or SME, five minutes gets you your own position against the region.
At a glance
Six things the data says
Across founders and business leaders in 16 Asia-Pacific markets, 86% use AI regularly or have it embedded across the business. Most spend little to do it: 74% are at $500 a month or less. ChatGPT (84%), Gemini (73%) and Claude (76%) lead the tools, 87% report time savings, and 90% plan to use more AI this year.
86%
use AI regularly or have it embedded
AI is part of daily work, not a parked experiment.
74%
spend $500 a month or less
Deep adoption on shoestring budgets. 41% are under $100.
84% / 73% / 76%
ChatGPT, Gemini, Claude
A three-horse race, not a one-tool region.
87%
see time savings; fewer see revenue
Only 21% report a revenue lift from AI so far.
50%
name cost as the top obstacle
Even though most spend little. 31% see no real obstacle at all.
90%
plan to increase AI use this year
62% significantly. Not one company plans to cut back.
Who responded
A founder-led, small-company sample
88% of respondents are founders or co-founders, and 74% run teams of 15 or fewer. Read this as a clear picture of early-stage and small-business AI use in Asia-Pacific, not of large enterprises. Six markets, Indonesia, Singapore, Bangladesh, Malaysia, Thailand and the Philippines, carry enough responses to stand on their own; the rest are part of the regional whole.
By market
By company size
By sector
Finding 01
Deep adoption, shoestring budgets
These companies have woven AI deep into how they work. 51% say it is embedded across multiple areas of the business, not parked in one experimental corner. Yet 74% pay $500 a month or less for it, and 41% spend under $100 or stay on free tiers entirely.
The contradiction is in what they call hard. Cost is the single most-cited obstacle at 50%, even though almost nobody is spending much. For this group AI is still a discretionary line to keep small, not an investment to scale up. At the same time, 31% report no significant obstacles at all.
Depth of AI use
- Embedded across the business51%
- Regular in one area34%
- Experimenting12%
- Aware, not started2%
- Not using, no plans1%
Monthly AI spend
- $0 (free tiers only)12%
- Under $100 / month30%
- $100–$500 / month35%
- $501–$2,000 / month15%
- Over $2,000 / month8%
Share of those who named a spend band.
Embedded adoption rises with team size
Share of each size band with AI embedded across the business. Larger small-teams adopt more deeply, not less.
Deep dive coming soon
Finding 02
The frontier is a three-horse race
ChatGPT still leads at 84%, but it is not running away with the region. Claude (76%) and Gemini (73%) sit right behind it, the three of them inside 11 points. In Asia-Pacific small companies, Google has caught up to the field rather than trailing it.
There is a gap lower down the stack. 72% use AI for software development, but only 33% reach for a dedicated coding tool like Copilot or Cursor. Most of that building still happens in a general chat window, not a purpose-built assistant.
By market, the picture shifts: Claude runs strongest in Thailand and Bangladesh, while ChatGPT reaches its regional high of 89% in Indonesia. The table shows the three leaders across every market with at least 13 responses.
Tools used in the past 30 days
Respondents could select more than one tool.
| Market | % of responses | ChatGPT | Gemini | Claude |
|---|---|---|---|---|
| Indonesia | 21% | 89% | 71% | 71% |
| Singapore | 16% | 86% | 72% | 76% |
| Bangladesh | 14% | 85% | 65% | 81% |
| Malaysia | 11% | 80% | 75% | 80% |
| Thailand | 9% | 76% | 71% | 82% |
| Philippines | 8% | 80% | 80% | 80% |
Markets shown carry 13 or more responses each. Smaller markets (Vietnam, India and others) point the same way but are too small to report as precise figures.
Deep dive coming soon
Prefer to read it later? Get the full report as a PDF — every chart, quote and country breakdown in one file.
Finding 03
Where AI shows up first
AI enters these companies through the front office before the back office. Content and marketing leads at 75%, with software development close behind at 72%. Customer support and recruiting sit at the bottom, the areas where trust and accuracy matter most and where founders are slowest to hand work to a model.
Respondents could select more than one area.
Finding 04
Productivity is here. Revenue is not, yet.
The wins these companies report are about getting work done faster. 87% point to time savings for the team and 79% to faster product development. Higher-quality output follows at 59%.
Top-line impact is a different story. Only 21% report an increase in revenue from AI so far. That is the classic early-curve pattern: efficiency lands first, and revenue is the harder, later prize. It also explains why so many founders frame AI as a cost rather than a growth lever.
But depth changes the maths. Companies with AI embedded across the business are nearly three times as likely to report a revenue lift as those who use it only in one area, and they pull ahead on every other result too.
Results seen from AI
Base: the 155 respondents who use AI regularly or have it embedded, the only ones asked this question. Respondents could select more than one result.
| Result | Embedded | Regular in one area |
|---|---|---|
| Significant time savings | 90% | 82% |
| Faster development | 84% | 71% |
| Higher quality work | 60% | 58% |
| Lower operating costs | 58% | 31% |
| Increase in revenue | 28% | 10% |
The depth dividend: going from regular use in one area to embedded across the business lifts reported cost savings from 31% to 58%, and makes a revenue lift nearly three times as likely.
Deep dive: AI is making Southeast Asia's startups faster, not richer, yet on e27, written from an earlier cut of this data
Finding 05
What is actually in the way
Cost tops the list at 50%, followed by data privacy and security at 28% and a team skills gap at 24%. Notably, 31% say there are no significant obstacles at all, the second most common answer.
The obstacles that rank lowest are telling. Integration friction and unclear ROI sit near the bottom, which fits a group that has already adopted: the hard part is no longer getting started, it is keeping up, staying accurate, and turning use into results.
Read together, the top two answers tell one story. The group that names cost as the biggest barrier is also the group spending the least, while nearly a third report no real barrier at all. The obstacle is rarely the tool itself. It is the decision to treat AI as a budget line worth growing.
50%
name cost as their biggest obstacle, yet 41% spend under $100 a month. The barrier is mindset, not price.
Biggest obstacles to adoption
Respondents could select more than one obstacle.
Finding 06
The direction is one-way
90% plan to increase their use of AI over the next 12 months, and 62% plan to increase it significantly. The remainder expect to hold steady. Not a single company in the sample plans to cut back.
For a group that already calls itself a deep adopter, this is the more striking number. They are not at a ceiling; they see headroom. The spending restraint of today is a starting point, not a settled budget.
12-month AI outlook
- Plan to increase significantly62%
- Plan to increase slightly28%
- Expect to stay the same10%
Benchmark yourself
How does your company compare?
Use this as a quick self-check. The middle column is the typical Asia-Pacific small company in this sample; the right column is the supporting figure. Pick your own answer in the last column, then turn it into a scorecard you can share.
| Dimension | The typical respondent | The figure | Your company |
|---|---|---|---|
| How deeply AI is used | Embedded across the business | 51% are at this depth; 86% use it regularly or deeper | |
| Monthly AI spend | $100–$500 a month | 74% spend $500 or less; 41% spend under $100 | |
| Primary tools | ChatGPT, Gemini and Claude | 84% / 73% / 76% | |
| Where AI is applied first | Content & marketing, then software | 75% content, 72% development | |
| Biggest result so far | Time savings for the team | 87% report it; only 21% report a revenue lift | |
| Most-cited obstacle | Cost of paid tools | 50% name cost; 31% see no real obstacle | |
| 12-month outlook | Increasing AI use | 90% plan to increase; 62% significantly |
Pick at least three answers above, then generate a shareable image.
In their words
The single biggest challenge, named
Swipe to read all 18
The biggest challenge by far is trust. AI still makes mistakes and hallucinations slip through, and we have not been able to hit 100% accuracy consistently. Every meaningful output still needs human review, which caps how much you can actually automate.
One challenge we have identified is the risk of overreliance on AI. Without proper guidance and continuous learning, team members may become too dependent on these tools and fail to develop the skills needed to think critically, solve problems independently, and grow professionally.
Honestly, we are maxing out the available hardware in our AI lab. We are running very cost-optimised on-prem Ollama instances, but our hardware outlay is budget.
The biggest challenge is that we have not used AI for generating revenue. Current usage is mostly to optimise internal operations. We have not found an AI tool we can use to help generate revenue in our current market.
The single biggest challenge is not adoption; most businesses are already using AI in some form. The real challenge is governance, ensuring that as AI capability scales, accountability scales with it. Most organisations have not solved that yet.
Our biggest challenge is bridging the gap between fragmented factory-level data and the high-integrity, structured data required for AI to generate verifiable Digital Product Passports that meet strict EU regulations.
Quoted with permission. Respondents who asked to stay anonymous are counted in the data but not named here. Lightly edited for clarity.
What it means
Three takeaways for operators
The starting line is behind us. The interesting question for Asia-Pacific founders is no longer whether to use AI; 86% already use it regularly or have it embedded. It is how deep to take it, because depth is what separates the companies seeing real results from the ones seeing only convenience.
Budget is not the constraint people think it is. Cost is the loudest complaint, yet only 47% of the companies naming it spend under $100 a month. The binding constraint is attention and skill, not licence fees. The companies pulling ahead are spending more time, not necessarily more money.
Revenue is the next frontier. Efficiency is banked; only 21% have turned AI into top-line growth. The deeper adopters are nearly three times as likely to get there. The move from a faster team to a bigger business is the work of the next 12 months, and 90% intend to keep climbing.
Interactive
Explore the data yourself
Filter the responses by country, sector, company size and growth stage, and watch every chart recompute in real time.
Based in Asia-Pacific? See how your own company compares, and your answers join the next round.
Filters
This filter leaves a small sample. Read the percentages as directional, not precise.
No founders match this combination of filters.
Based on n = 181 responses. Percentages are of the currently filtered group. Single-select questions (spend, outlook) sum to 100%; multi-select questions (tools, areas, results, obstacles) do not.
Founder or business leader in Asia-Pacific?
See how your company compares
Answer the same questions these 181 companies did and you land straight on your own comparison: how deep your AI use runs, what you spend, and what you are getting back against the regional benchmark. About five minutes, and your answers sharpen the next update of this data.
Get my benchmarkKeep a copy
Get the PDF of this report
Drop your email and I will send you the full report as a PDF to read, share or forward to your team.
Questions, answered
Asia-Pacific AI adoption, in brief
How many Asia-Pacific small businesses use AI in 2026?
Across founders and business leaders in 16 Asia-Pacific markets, 86% use AI regularly or have it embedded across the business, and 51% have it embedded across multiple areas. Only a small minority are still experimenting or have not started.
How much do Asia-Pacific founders spend on AI each month?
Most run lean. 74% spend $500 a month or less on AI, and 41% spend under $100. Deep adoption does not require a large budget; the binding constraint is time and skill, not licence fees.
Which AI tools do Southeast Asian and Asia-Pacific businesses use most?
ChatGPT leads at 84%, followed by Claude at 76% and Gemini at 73%. GitHub Copilot or Cursor reach 33%. The frontier is a three-horse race rather than a single-tool region.
Where do small businesses use AI first?
Content and marketing comes first at 75%, followed by software development at 72%. Product research, internal operations and sales follow. AI shows up earliest in the work every small company already does daily.
What results are companies actually getting from AI?
87% report significant time savings for the team and 79% report faster product development. Revenue is the laggard: only 21% report a revenue lift from AI so far. Efficiency is banked; top-line growth is the next frontier.
What is the biggest obstacle to AI adoption for small businesses?
Cost is the most-named obstacle at 50%, even though most of this group spends little. 31% report no significant obstacle at all. Team skills gaps and output reliability are the next concerns after cost.
Are Asia-Pacific businesses planning to use more AI?
Yes. 90% plan to increase their AI use over the next 12 months, with 62% planning a significant increase. Not one company in the sample plans to cut back; the direction across Asia-Pacific is one-way.
What is AI adoption?
AI adoption is the extent to which a company actually uses AI tools in its everyday work, from one-off experiments to AI embedded across multiple business areas. In this survey, 86% of Asia-Pacific founders use AI regularly or have it embedded.
Method & thanks
How this survey was run
The figures in this report (n = 181) come from founders and business leaders across 16 Asia-Pacific markets, fielded 13 April 2026 to 1 September 2026. The sample skews to small, founder-led companies: 88% are founders or co-founders and 74% run teams of 15 or fewer. Read it as a clear picture of early-stage and small-business AI use, not of large enterprises.
Responses were de-duplicated by email, keeping the most recent submission per person, and out-of-region and test entries were removed. Single-select questions are reported as percentages that sum to 100; multi-select questions (tools, areas, results, obstacles) allow more than one answer and do not. Where a market or segment carries fewer than 13 responses, it is described as directional and never presented as a precise headline figure. Free-text answers are stripped of any links or contact details before publishing, and names appear only where a respondent gave explicit permission to be quoted. All dollar amounts in this report are in US dollars (USD).
Huge thanks to
for sharing the survey with their founder community, and to every founder who answered honestly.
Cite as: Dhawal Shah, APAC AI Adoption 2026: The Full Report, dhawalshah.net, 1 June 2026, updated 1 September 2026. Free to quote and link with attribution.