No platform has won SEA
81.7%Builders use more than one AI platform, making multi-homing the default.
2026 Report Preview
State of AI Builders in Southeast Asia
Evidence from 2,719 approved builders across 55 countries, drawn from Agentic AI Build Week 2026 registration data.
Top Takeaways
What the preview reveals
The full report carries the deeper segmentation. The public preview gives readers the core evidence: scale, technical depth, platform behavior, and why the builder community matters now.
Claude closes the gap
76.6%Claude sits near OpenAI overall and strengthens among experienced builders.
The pipeline is young
74.7%Nearly three quarters of respondents are in their first two years of AI/ML.
Vietnam anchors the map
64.4%The host-city effect is real, but the composition points to structural depth.
Chinese models are meaningful
27.5%More than one in four builders use at least one Chinese AI model.
Trust networks drive growth
~1/3Referral counters attribute roughly a third of the community to connectors.
Dataset and Caveats
Methodology snapshot
These findings describe AABW registrants: an engaged, self-selected builder community. They should be read as a view of the leading edge, not a representative estimate of the entire Southeast Asian AI workforce.
Chapter Lineup
What readers get from the complete report
The free preview establishes credibility. The locked chapters carry the deeper segmentation, implications, and decision guidance.
01Preview
Builder Map
Geography, seniority, roles, and the concentration of AI talent.
Preview insight02Preview
Platform Race
OpenAI, Claude, Gemini, Chinese models, cloud access, and tooling.
Preview insight03Full report
Experience Curves
How early-career and senior builders adopt different stacks.
Included in the complete 2026 analysisRequest access04Full report
Market Profiles
Country-level differences across Vietnam, Singapore, India, and more.
Included in the complete 2026 analysisRequest access05Full report
Implications
What enterprises, policymakers, platforms, and investors should do next.
Included in the complete 2026 analysisRequest access06Full report
Recommendations
Priority actions for ecosystem builders and institutional partners.
Included in the complete 2026 analysisRequest accessPublic Preview Findings
Three signals strong enough to publish free
Each figure is paired with one interpretation, matching the Stanford AI Index habit of making the chart do real evidentiary work.
Figure 1
Country share of approved builders
Vietnam is the anchor, but the signal is wider than geography.
Source: AABW registration data, 2026 | Chart: GenAI Fund
Interpretation
Vietnam is the anchor, but the signal is wider than geography.
Vietnam accounts for 64.4% of approved builders, partly reflecting the Ho Chi Minh City location. The U.S., Singapore, India, and Malaysia form the next tier, pointing to diaspora reach and regional participation.
Figure 2
Platform adoption among builders
The platform race is a portfolio race.
Source: AABW profile survey, 2026 | Multi-select field
Interpretation
The platform race is a portfolio race.
OpenAI and Claude sit in near-parity, while Gemini remains a major third platform. Adoption sums beyond 100% by design because builders use multiple tools at once.
Figure 3
Project focus areas
The leading project categories are enterprise-relevant.
Source: Keyword analysis of free-text project descriptions | Directional
Interpretation
The leading project categories are enterprise-relevant.
Agentic systems, automation, conversational AI, and retrieval systems are not decorative experiments. They map to operational workflows that enterprises are actively trying to redesign.
Full Report Access
The preview ends where the segmentation gets valuable.
Unlock the full report for seniority curves, country spotlights, platform segmentation, trust-network analysis, market implications, and recommendations.
Premium Sections
The full report explains what the averages hide
Deeper segmentation shows how experience, country, role, and organisation type change the meaning of the aggregate platform race.
Figure 4
Platform adoption by category and experience
Source: AABW profile survey, 2026 | Structured tool-selection field
Locked table previewExperience x platform category
Figure 5
Model adoption by market
Source: AABW profile survey, 2026 | Selected major markets
MarketOpenAIClaudeGeminiChinese model
Vietnam78%76%63%26%
Singapore86%86%58%38%
India82%81%61%29%
Malaysia75%64%58%30%
Philippines80%71%66%26%
Indonesia82%73%79%42%
Locked interpretation
Markets differ in adoption logic.
Singapore shows high Claude adoption and strong Chinese model experimentation. Indonesia is small but striking: high seniority, high Gemini usage, and the highest Chinese model adoption among major SEA markets.
For enterprises
Map the work before deploying tools. Create role-specific training and AI-translator roles.
For governments
Treat AI talent as a workforce system, not only a technical training pipeline.
For platforms
Invest in the evaluation layer and early-career cohorts before loyalties harden.
Final Signal