OpenAI has released country-level data on ChatGPT usage for the first time. In the accompanying Signals dataset, Singapore ranks first among 147 countries in pe…
Read moreAISG launched in 2017 with up to S$150M over five years from NRF, over S$500M cumulatively. It runs research, talent, products and governance at once: SEA-LION covers 11 Southeast Asian languages, AIAP has graduated about 500–600 apprentices across 22 cohorts, and NOAI has taken four golds across two IOAI editions. The piece also clears up two names often misattributed to AISG (AI Trailblazers, Kampong AI) and the bottleneck after 100E was archived: AIAP capacity of about 60 apprentices per cohort.
New /ask/ page: multi-turn chat grounded in the site’s debates, policies, tracker and other datasets, with source links, available in all five languages. Entry points in the homepage sidebar and the navigation.
The site's first annual synthesis: the tracker's six dimensions (investment, talent, compute, adoption, research, governance) on one page — S$139 per-capita government AI spend, 5,000 of 15,000 AI practitioners, 1.4 GW of compute, 62.5% large-enterprise vs 14.5% SME adoption, per-capita papers #1, and rule-maker status. Three judgments: the amplification ratio (roughly S$13 of hyperscaler commitments per S$1 of government spend) is the real metric; the SME gap is the main battlefield for the next 2–3 years; rule-making is not rule-enforcement. No new numbers — every figure carries its own as-of date and source.
Why We Hire AIAP Graduates — Heineken.
Theseus Infrastructure; Razer.
Temasek's annual review, published 2026-07-08, gives AI its own chapter (§3.3 Our AI Strategy). It carries exactly one dated numeric commitment: AI-related investments now make up 6% of the portfolio, to be raised to 10–15% by 31 March 2031 — against a S$518 billion portfolio, reaching 15% means more than doubling AI holdings. The piece walks through the four pillars, the year's purchases (Anthropic / OpenAI / xAI / CuspAI / PhysicsX / Lam Research and others), and what lands in Singapore: Temus's AI Foundry, Aicadium × gategroup, Resaro, and the AI Leadership Programme run with Stanford Graduate School of Business.
On 2026-08-06 OpenAI published country-by-country ChatGPT usage data for the first time. The accompanying Signals dataset ranks Singapore #1 of 147 countries on messages per capita for Q2 2026 (UAE #6, United States #51, Japan #57, Indonesia #101). This is the third independent vendor dataset to place Singapore at or near the top, after Anthropic (AUI 5.53, 1st of 116 regions) and Microsoft (60.9% adoption, 2nd globally). Singapore appears in the article body only as a counter-example — the share of messages from users aged 35+ rose just 1.2 percentage points against a 5.1-point country median; our editor's note explains why that figure measures message composition rather than adoption.
Resaro; Temus.
Every parliamentary question, response, and debate on AI from 2015 to 2026.
12 continuously updated datasets — each card shows its latest addition.
Four National AI Missions
The first National AI Mission — AI-driven robotics, factory digitalisation and the upgrading of manufacturing.
One of the National AI Missions — AI automating airports, ports and logistics to strengthen Singapore’s global-hub position.
MAS's AI supervision and financial-sector adoption — risk-management guidelines, fintech and AI inside financial institutions.
AI in health — from the national health mission to hospital systems and healthtech companies.
Understand the framework
The state wraps itself around enterprise AI-native transformation. The state itself doesn’t need to become AI-native — it just needs to amplify how fast enterprises do.
Different questions, different entry points. Tell us what you’re trying to do.
Six-lever map, policy evolution, international benchmarks — the whole story, joined up.
Read the core argument →Policy library, AI legal framework, parliamentary debates, references — citable sources.
Browse policy & debates →Startup ecosystem, ecosystem map, talent pipeline, open source, international benchmarks.
Browse ecosystem →Recent updates, tracker readings, latest parliamentary debates, fresh policies.
View updates feed →Six levers, 115 concrete projects. Re-sliced by "how AI gets pushed in" — cutting across ministries into a single execution pipeline.
"AI-native isn’t scale — it’s architecture. A country cannot be AI-native "on its own"; it has to transmit through to enterprises."
Read the "AI-native country" essay