This page preserves a reference summary of Strategy 2: Advancing AI Research & Development from the National AI Roadmap 2021–2025, published by MOSTI (ISBN 978-967-19025-5-4) and reproduced for research and citation with attribution. The full text is in the Playbook PDF; the official copy is at mastic.mosti.gov.my. Independent four-year analysis of how this strategy fared is in Roadmap vs Reality.
If Strategy 1 was about who governs AI, Strategy 2 was about who creates it. The roadmap argued that Malaysia could not become a high-technology nation by importing AI alone — it needed indigenous research capacity, a pipeline of researchers, and a sustained lift in national research intensity. Strategy 2 set out to move Malaysia from being a deployer of AI toward becoming a developer of it.
The diagnosis was blunt: Malaysia's gross expenditure on R&D (GERD) sat at roughly 1.08% of GDP, far below the levels of the advanced economies it hoped to join, and AI-specific research was thin and scattered. Strategy 2 proposed to concentrate effort on priority domains, fund the people and institutions to pursue them, and ring-fence money so AI research did not get crowded out.
What Strategy 2 set out to do
The strategy spanned the full research value chain — from setting priorities, to funding talent and centres, to providing the compute and the money to sustain them.
- Identify and continuously update AI R&D priority areas, and pursue next-generation AI systems aligned to national needs rather than chasing every frontier.
- Strengthen AI research human capital — postgraduate scholarships, post-doctoral fellowships, and funding for AI-based research projects, including in non-science domains.
- Establish AI Centres of Excellence (AI-XL) within research institutions to anchor sustained, specialised research capacity.
- Provide a cloud-based policy sandbox giving researchers access to hyperscale compute for AI R&D without each institution building its own.
- Ring-fence dedicated AI R&D funding and lift national research intensity (GERD) over the decade.
Targets at a glance
| Target | Horizon |
|---|---|
| Ring-fence RM10m, then a further RM15m, for AI R&D in priority areas | H1–H2 |
| 20 postgraduate scholarships (Master’s by research & PhD) | H1 |
| 20 post-doctoral fellowships in AI | H1 |
| 50 AI-based non-science & technology research projects funded | H1 |
| 10 AI Centres of Excellence (AI-XL) established | H1–H2 |
| 20% of relevant allocations directed to AI R&D investments | H2–H3 |
| Lift GERD from 1.08% toward 3.5% of GDP | by 2030 |
Why R&D intensity mattered to the model
The roadmap's economic case — that AI could add roughly 1.2 percentage points to Malaysia's ~4.4% baseline growth, a ~30% uplift — rested on Malaysia building genuine capability, not just buying tools. Research intensity was the lever: without a step-change in GERD and a domestic research base, the productivity gains AI promised would accrue mainly to the foreign firms supplying the technology.
The roadmap ring-fenced tens of millions of ringgit for AI research. No consolidated public reporting against those R&D targets or the GERD trajectory was ever published; larger AI research allocations that appeared later under Budget 2025 belong to the National AI Office era, not to this roadmap. See how Strategy 2’s targets were tracked →