Start with the people.
73 of 129 respondents cite a lack of AI skills or knowledge. Capability needs time, training and experience to develop.
THE PHILIPPINE
AI REPORT.
The ambition is here.
Now comes the hard part.
AI has entered the Philippine workplace. Turning that momentum into lasting capability is the next chapter.
Explore the findings ↓RESEARCH BY SWARMTIM SANTOS · PIA BESMONTE LIGOT-GORDON
PHILIPPINE ARCHIPELAGO / HOVER TO REVEAL
175organizations surveyed
47research questions
2025survey year
A snapshot of enterprise AI.
A signal of what comes next.
AI adoption has momentum.
What will turn it into lasting capability?
175 organizations. 47 questions.17 figures, one unfolding story ↓
01
Who is
speaking?
Before reading the findings, meet the sample behind them. The survey brings together 175 organizations across industries and sizes, with a strong technology presence.
organizations in the survey
Responses came through professional networks, industry associations and direct outreach. This is a view into participating organizations, not a nationally representative estimate.
Who answered the survey?
n = 175 · Single selection
ONE PERSON ICON = ONE RESPONDENT · n = 175
C-suite / senior executive70 · 40.00%
Upper management24 · 13.71%
Manager / team lead37 · 21.14%
Supervisor1 · 0.57%
Individual contributor37 · 21.14%
Other6 · 3.43%
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| C-suite / senior executive | 40.00% | 70 |
| Upper management | 13.71% | 24 |
| Manager / team lead | 21.14% | 37 |
| Supervisor | 0.57% | 1 |
| Individual contributor | 21.14% | 37 |
| Other | 3.43% | 6 |
Which industries are represented?
n = 175 · Single selection
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Technology / software / IT services | 37.14% | 65 |
| Financial services / banking / insurance | 13.71% | 24 |
| Professional services / consulting | 6.29% | 11 |
| Research / academia | 4.57% | 8 |
| Retail / consumer goods / eCommerce | 4.57% | 8 |
| Energy / utilities / oil & gas | 4.00% | 7 |
| Transportation / logistics / supply chain | 3.43% | 6 |
| Healthcare / pharmaceuticals / life sciences | 2.86% | 5 |
| Media / entertainment / publishing | 2.86% | 5 |
| Education / EdTech | 2.29% | 4 |
| Government / public sector | 2.29% | 4 |
| Nonprofit / NGOs / social impact | 2.29% | 4 |
| Hospitality / travel / leisure | 1.71% | 3 |
| Manufacturing / industrial goods | 1.71% | 3 |
| Real estate / construction | 1.71% | 3 |
| Telecommunications | 1.71% | 3 |
| Business process outsourcing | 1.14% | 2 |
| Legal services / law firms | 0.57% | 1 |
| Other | 5.14% | 9 |
How large are the organizations?
n = 175 · Single selection
Fewer than 100 employees54.86%
100–999 employees14.29%
1,000–4,999 employees8.57%
5,000–9,999 employees9.71%
10,000+ employees12.57%
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Fewer than 100 employees | 54.86% | 96 |
| 100–999 employees | 14.29% | 25 |
| 1,000–4,999 employees | 8.57% | 15 |
| 5,000–9,999 employees | 9.71% | 17 |
| 10,000+ employees | 12.57% | 22 |
What kinds of organizations took part?
n = 175 · Single selection
Early-stage startup (pre-seed / seed)16.00%
Venture-backed startup (Series A+)10.29%
Mid-market / SME18.29%
Local enterprise17.71%
Multinational enterprise16.00%
Government & public sector5.14%
Other16.57%
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Early-stage startup (pre-seed / seed) | 16.00% | 28 |
| Venture-backed startup (Series A+) | 10.29% | 18 |
| Mid-market / SME | 18.29% | 32 |
| Local enterprise | 17.71% | 31 |
| Multinational enterprise | 16.00% | 28 |
| Government & public sector | 5.14% | 9 |
| Other | 16.57% | 29 |
Read what follows with those perspectives in mind. The first finding is how far everyday use has already spread.
02
Already part
of the working day.
For most respondents, AI is already in use. The more useful question is where it fits into work, and how established that habit has become.
%
reported some AI use in the past year.
104 of 113 responses · 92.04% · p. 71ONE SQUARE = ONE RESPONSE
■ SOME AI USE □ NO AI USE
Where was AI used in 2025?
n = 113 · Multiple selections allowed
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Internal task automation | 64.60% | 73 |
| Data analysis / decision-making | 60.18% | 68 |
| Customer-service chatbots | 41.59% | 47 |
| Content creation | 63.72% | 72 |
| Recruitment / HR | 23.01% | 26 |
| Personalization / recommendations | 38.94% | 44 |
| Forecasting / prediction | 36.28% | 41 |
| Pilot / test projects | 45.13% | 51 |
| Employee AI training | 27.43% | 31 |
| No AI usage in the past year | 7.96% | 9 |
How long have people used generative AI?
n = 118 · Single selection
Less than 1 month1.69%
1–6 months23.73%
7–12 months19.49%
More than 12 months54.24%
Not currently using AI tools0.85%
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Less than 1 month | 1.69% | 2 |
| 1–6 months | 23.73% | 28 |
| 7–12 months | 19.49% | 23 |
| More than 12 months | 54.24% | 64 |
| Not currently using AI tools | 0.85% | 1 |
03
Tools are widespread.
The work goes deeper.
Chat interfaces make powerful AI easy to access. Building and sustaining AI inside an organization calls for a different mix of tools, projects and expertise.
%
ANDuse ChatGPT / OpenAI
%
use PyTorch / TensorFlow / JAX
97 and 14 of 117 respondents respectively. Tool selections overlap; these are not mutually exclusive groups.
Which AI tools are used at work?
n = 117 · Multiple selections allowed
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| ChatGPT / OpenAI | 82.91% | 97 |
| Gemini / Google | 62.39% | 73 |
| Claude / Anthropic | 43.59% | 51 |
| Microsoft Copilot / Azure OpenAI | 39.32% | 46 |
| GitHub Copilot | 28.21% | 33 |
| Canva Magic Write / AI | 23.93% | 28 |
| DALL·E | 14.53% | 17 |
| Notion AI | 12.82% | 15 |
| GrammarlyGO | 11.97% | 14 |
| Midjourney | 11.97% | 14 |
| PyTorch / TensorFlow / JAX | 11.97% | 14 |
| CUDA | 10.26% | 12 |
| Jasper | 4.27% | 5 |
| Adobe Firefly / Sensei | 3.42% | 4 |
| Salesforce Einstein GPT | 2.56% | 3 |
| Writer | 1.71% | 2 |
| Not currently using AI tools | 1.71% | 2 |
| Other | 15.38% | 18 |
What kinds of AI projects are underway?
n = 107 · Multiple selections allowed
Source note
The p. 10 image conflicts with the detailed table: vendor evaluation 26% vs 27.10%; third-party integration 24% vs 25.23%; internal tooling 40% vs 36.45%; applied development 36% vs 29.91%; application development 34% vs 46.73%; end-user enablement 13% vs 41.12%. This reconstruction uses p. 70. Open the original to compare.
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Proof of concept | 65.42% | 70 |
| Vendor evaluation | 27.10% | 29 |
| Large-scale integration of third-party AI | 25.23% | 27 |
| Internal AI tooling development | 36.45% | 39 |
| Fundamental AI research | 28.04% | 30 |
| Applied AI development | 29.91% | 32 |
| AI application development | 46.73% | 50 |
| End-user AI enablement | 41.12% | 44 |
04
A mandate needs
a structure.
Senior leaders are driving AI strategy in much of the sample. The presence of specialist roles and views on leadership add further context to that commitment.
Who leads AI strategy?
n = 175 · Single selection
Executives60.57%
IT teams14.86%
Business teams10.29%
Employees / users2.29%
No one / individual efforts12.00%
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Executives | 60.57% | 106 |
| IT teams | 14.86% | 26 |
| Business teams | 10.29% | 18 |
| Employees / users | 2.29% | 4 |
| No one / individual efforts | 12.00% | 21 |
Which AI roles exist inside companies?
n = 66 · Multiple selections allowed
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| AI engineer | 30.30% | 20 |
| Data scientist | 36.36% | 24 |
| Software developer with AI work | 33.33% | 22 |
| Machine learning specialist | 13.64% | 9 |
| Data engineer | 30.30% | 20 |
| AI product manager | 10.61% | 7 |
| Data analyst using AI | 19.70% | 13 |
| AI strategy lead | 16.67% | 11 |
| Business intelligence analyst | 21.21% | 14 |
| AI project manager | 15.15% | 10 |
| AI trainer / data labeling | 3.03% | 2 |
| Prompt engineer | 7.58% | 5 |
| UX designer for AI tools | 10.61% | 7 |
| AI compliance / governance officer | 12.12% | 8 |
| Research scientist / engineer | 9.09% | 6 |
| No AI-related roles | 19.70% | 13 |
| Unaware of AI-related roles | 9.09% | 6 |
| Other | 3.03% | 2 |
How is leadership’s approach received?
n = 130 · Single selection
← SATISFIEDDISSATISFIED →
NEUTRAL RESPONSES STRADDLE THE CENTER
Very satisfied20.00%
Somewhat satisfied39.23%
Neutral25.38%
Somewhat dissatisfied8.46%
Very dissatisfied6.92%
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Very satisfied | 20.00% | 26 |
| Somewhat satisfied | 39.23% | 51 |
| Neutral | 25.38% | 33 |
| Somewhat dissatisfied | 8.46% | 11 |
| Very dissatisfied | 6.92% | 9 |
05
Ambition meets
organizational reality.
The most frequently cited barrier is a lack of AI skills or knowledge. Security concerns and unrealistic expectations follow. The work ahead spans people, trust and implementation.
06
More room
for human work.
Respondents describe more time to think, quicker decisions and less writing. The reported benefits sit alongside a more varied picture of workforce effects.
%
More time for strategic thinking.
102 of 134 respondents selected this benefit. These are self-reported experiences, not measured productivity gains.
How has generative AI helped at work?
n = 134 · Multiple selections allowed
Source note
The p. 17 image labels “no noticeable change” as 1%; the appendix gives 3 of 134 responses, or 2.24%. This chart follows the appendix.
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| More time for strategic thinking | 76.12% | 102 |
| More time for innovation | 58.21% | 78 |
| Quicker decisions | 65.67% | 88 |
| Better collaboration | 43.28% | 58 |
| Less time writing | 67.91% | 91 |
| Fewer administrative tasks | 51.49% | 69 |
| No noticeable change | 2.24% | 3 |
What headcount loss was linked to AI?
n = 119 · Single selection
ONE SQUARE = ONE RESPONSE · n = 119
Zero layoffs83 · 69.75%
1–10011 · 9.24%
100–1,0003 · 2.52%
1,000+2 · 1.68%
Unaware of AI-related layoffs17 · 14.29%
Other3 · 2.52%
Source note
“Unaware of layoffs” is not confirmation of zero layoffs. The source category boundaries overlap at 100 and 1,000; labels are retained as published.
Exact data & original figure +
| Response | Share | Count |
|---|---|---|
| Zero layoffs | 69.75% | 83 |
| 1–100 | 9.24% | 11 |
| 100–1,000 | 2.52% | 3 |
| 1,000+ | 1.68% | 2 |
| Unaware of AI-related layoffs | 14.29% | 17 |
| Other | 2.52% | 3 |
07
The next step
is an intention.
The report’s 2025 survey asked about plans for 2026. HR and employee training show the largest planned increases, followed by forecasting and customer service.
LARGEST PLANNED INCREASE+pp
Recruitment / HR
23% usage → 43% planned
2025 usage and 2026 plans
Intent reported in 2025 · Rounded comparison values
Source note
Intent reported in 2025, not observed outcomes. The final row is inconsistent: the table prints 8% → 6% with a −7 change, while the prose says 2%. We retain the printed endpoints, show their arithmetic difference (−2 pp), and flag the conflict. The planned-use question’s response base is not supplied in the appendix.
Exact data & original figure +
| Response | 2025 | 2026 plans |
|---|---|---|
| Recruitment / HR | 23% | 43% |
| Employee AI training | 27% | 43% |
| Forecasting / prediction | 36% | 51% |
| Customer service | 42% | 57% |
| Personalization | 39% | 48% |
| Data analysis / decisions | 60% | 64% |
| Internal automation | 65% | 68% |
| Content creation | 64% | 64% |
| Pilot / test projects | 45% | 44% |
| No AI usage / plans | 8% | 6% |
Intent points to the next set of use cases. Delivering them brings the story back to skills, governance and the capacity to build.
08
Build knowledge.
Broaden capacity.
The report also looks beyond workplace adoption to AI research. These figures measure publications, offering a separate view of the institutions and regional context behind the talent pipeline.
AI research output across ASEAN
Scopus-indexed publications · Research contribution
APPROXIMATE VALUES · RECONSTRUCTED FROM THE REPORT’S PLOTTED MARKS
Publications, approximately
All ten countries are represented. Smaller series are approximated by matching the source legend colors. Myanmar’s visible cyan fragments are connected with an approximate line; hidden annual values remain unavailable. Use the 0–250 view to inspect the smaller traces. 2025 reflects the search as of November and may be incomplete.
Reconstruction notes & original +
Approximate raster digitization of PDF pages 44–45. Not the underlying Scopus dataset. Line values rounded to 25 publications; bars to 10. Occluded points remain null. Small differences should not be interpreted as exact rankings or changes.
Annual lines use the visible 2015–2025 axis, despite the original title saying 2021–2025. Small-series traces are matched to legend colors. Myanmar’s line connects surviving color fragments with straight segments. These connections are visual estimates, not observed annual values; no extrapolation beyond the recovered span is shown. Values near zero are below the raster resolution (about 13 publications per pixel).
Compare with original report figure ↗Where Philippine AI research is produced
Scopus-indexed publications · Research contribution
APPROXIMATE VALUES · RECONSTRUCTED FROM THE REPORT’S PLOTTED MARKS
01De La Salle University≈1,060
02Mapua University≈930
03University of the Philippines Diliman≈480
04National University, Philippines≈340
05Ateneo de Manila University≈270
06University of the Philippines, Manila≈170
07FEU Institute of Technology≈170
08Technology Institute of the Philippines, Quezon City≈170
09Technological Institute of the Philippines, Manila≈150
10Polytechnic University of the Philippines≈140
11University of Santo Tomas, Manila≈140
12Batangas State University≈140
13Technological University of the Philippines≈130
14University of the Philippines, Los Baños≈120
15University of the Cordilleras≈120
16Isabela State University≈100
17Adamson University≈100
18Mindanao State University, Iligan Institute of Technology≈90
19Cebu Technological University≈90
20Lyceum of the Philippines University, Manila≈80
21Caraga State University≈80
22Angeles University Foundation≈80
23Cavite State University≈70
24University of San Carlos≈70
25José Rizal University≈70
26University of the Philippines, Open University≈70
27Cebu Institute of Technology University≈60
28Southern Luzon State University≈60
29University of Perpetual Help System DALTA≈60
30St. Paul University Philippines≈60
31Mapua Malayan Colleges Laguna≈50
32Pamantasan ng Lungsod ng Maynila≈50
33Bohol Island State University≈50
Reconstruction notes & original +
Approximate raster digitization of PDF pages 44–45. Not the underlying Scopus dataset. Line values rounded to 25 publications; bars to 10. Occluded points remain null. Small differences should not be interpreted as exact rankings or changes.
Bar lengths are calibrated against the printed 0–1,200 axis. Counts are rounded to 10 publications; closely spaced bars can round to the same value. The source order and all 33 affiliations are retained.
Compare with original report figure ↗09
Build the capacity.
Then build the scale.
The evidence points to a practical agenda: invest in people, make responsible use operational, and judge deployment against a defined business outcome.
Invest in people and processes.
Give teams structured learning time. Pair technical work with product and governance expertise. Build on the experience already accumulating inside organizations.
Return to the skills gap ↗Make responsible use operational.
Translate principles into data rules, ownership and evaluation. Give executive sponsorship the specialist support needed to guide implementation.
Return to specialist roles ↗Define success before the pilot.
Start with a specific business problem and a baseline. Evaluate what changes, then expand when the evidence supports the next step.
Return to project types ↗Figure index 17 exhibits · jump to any finding +
Research is
a collective effort.
The 15 researchers, writers, designers, analysts and contributors credited in the report. Names, portraits and roles as published in the 2025 edition.
Tim Santos
Project Lead
REPORT P. 79 ↗
Pia Besmonte Ligot-Gordon
Lead Writer
REPORT P. 79 ↗
Lennon Villanueva
Lead Designer
REPORT P. 79 ↗
Karl Ezra Pilario, PhD
Researcher & Writer
REPORT P. 79 ↗
Justine Rex Llanes
Researcher & Writer
REPORT P. 79 ↗
Kir Peñalber
Designer & Developer
REPORT P. 79 ↗
Alexis Collado
Producer
REPORT P. 79 ↗
Dexter Ligot-Gordon
Contributor
REPORT P. 79 ↗
Luis Gabriel Villadarez
Data Analyst
REPORT P. 80 ↗
Ashley Ko
Events Lead and Designer
REPORT P. 80 ↗
Tiffany Lim
Contributor
REPORT P. 80 ↗
Aldo Santos
Contributor
REPORT P. 80 ↗
Sherwin Pelayo
Contributor
REPORT P. 80 ↗
Dominic Ligot
Contributor
REPORT P. 80 ↗
Thea Montgomerie Anderson
Editorial Consultant
REPORT P. 80 ↗A useful signal.
A specific sample.
An online survey conducted in October–November 2025, distributed through professional networks, industry associations and direct outreach.
Based on The Philippine AI Report 2025, inaugural comprehensive version. Research by Swarm. Tim Santos, project lead; Pia Besmonte Ligot-Gordon, lead writer.
175organizations in the full survey
37%technology, software and IT services
55%have fewer than 100 employees
Question-level response counts vary. Percentages are rounded. Multiple-choice answers can overlap. Results are self-reported and are not nationally representative estimates. The survey did not collect respondent locations; the header map is illustrative.
Methodology and limitations: pp. 66–69. Detailed results: pp. 70–75. SOURCE P. 69 ↗