The roadmap was unusually specific about what success looked like: 875,000 MSMEs on cloud, 200,000 "future AI talents," 87,500 educators trained, 11 AI consortia, a top-40 Global Open Data Index ranking, RM10 million ring-fenced for R&D, and dozens more. It also specified how progress would be tracked — through the AI-CIU, KPIs for every initiative, annual economic-impact assessments, and a mid-term evaluation in 2023.
None of that tracking was ever published, as far as we can find. The targets below are therefore rated not as "hit" or "missed" — we cannot know without the data — but by the most honest categories the evidence supports: superseded by a later body, delivered through other means (a different programme or private capital achieved a similar end), or simply no public data. The last category dominates, and that is the story this scorecard tells.
"No public data" is not an accusation that a target failed — some may well have progressed. It means the government published nothing that lets a citizen check. A roadmap with KPIs but no public scorecard cannot be held to account, and tellingly, the successor 2026–2030 plan now promises an "annual National AI Report" — an admission, in effect, that the last plan had none.
44 targets, by status
Roughly four in five measurable targets have no public progress record at all.
The full scorecard
Filter by strategy or by status. Targets are quoted or closely paraphrased from the roadmap's strategic-initiative tables.
The accountability gap
This is not a story of a government that tried and failed — it may be that, but the public record does not let anyone say so. It is a story of a plan that set 44 measurable promises and then published nothing that allows those promises to be checked. The mechanisms for accountability existed on paper: the AI-CIU, the KPIs, the annual assessments, the 2023 mid-term review. They simply never produced visible output.
Where outcomes are visible, they largely arrived by other routes — AI skilling through AIForMYFuture, compute through private data centres, connectivity through JENDELA — none of them owned by this roadmap. The clearest verdict the evidence supports is the quietest one: the AI-Rmap was not so much failed as abandoned, and its replacement now promises the transparency it never had.
Targets are drawn from the strategic-initiative tables of the National AI Roadmap 2021–2025 (MOSTI; archived here), quoted or closely paraphrased. We searched official channels (MOSTI, the Ministry of Digital, NAIO/MyDIGITAL), budget documents and credible reporting for published progress against each KPI.
"Superseded" denotes a target whose function was taken over by a later body or programme (chiefly NAIO, est. Dec 2024). "Delivered via other means" denotes an outcome demonstrably achieved through a different vehicle (e.g. AIForMYFuture, private data-centre investment, JENDELA). "No public data" means no progress reporting against the original KPI could be located — not a determination of failure. Full sources: airmap.my/sources. Independent of MOSTI, the Ministry of Digital and NAIO.