The Singapore AI Pulse

The Pulse at a Glance

Eight numbers that frame where Singapore sits on AI, as of mid-2026. Everything else in this report expands on these.

Sources: MOM AI Adoption Among Firms report (Apr 2026), McKinsey × EDB × Tech in Asia, Smart Nation NAIS, Temasek Review 2026.

The National Push

Singapore moved from strategy to delivery in 2026. The refresh at ATxSummit was framed by Minister Josephine Teo as a "double-click rather than a system reboot", and the new machinery is sector-targeted, funded, and chaired at the top.

Four National AI Missions

Advanced Manufacturing, Financial Services, Connectivity and Healthcare. Together these sectors make up over 40% of GDP, and the government says the missions will be driven by "problem statements worth solving, not just for Singapore but for the world". Each mission pairs government enablers (data access, regulatory sandboxes) with private partners. Teo's ATxSummit keynote

New institutions, new money

Sources: MDDI NAIS refresh factsheet, Smart Nation, sgai policy profile, EDB on Kampong AI.

Partnerships with the frontier labs

Google and OpenAI have each signed MoUs with MDDI to co-develop AI across public and private sectors. NVIDIA picked Singapore for its second Asia-Pacific AI research lab, focused on embodied AI and efficient AI computing. The message from Teo was explicit: global leaders anchor here because of the network effect and the record of trusted adoption. EDB

The Adoption Reality

The MOM inaugural survey (fieldwork Jan–Mar 2026, 2,560 establishments, 486,600 workers) is the most honest read of the ground: AI is early-stage, uneven, and currently complementing rather than replacing labour.

Sector leaders: Information & Communications 74.1%, Professional Services 57.5%, Financial & Insurance 56.4%. Digitally intensive, knowledge-based sectors are pulling ahead while most of the economy lags.

Depth is shallow. Of the 28.5% who have adopted, only 3.8% have integrated AI into core business processes; 7.4% are still planning, 6.0% piloting. Singapore also trails Denmark, Finland and Sweden on firm-level adoption, and China's generative AI adoption stands at 42.8% by comparison.

Labour impact so far: only 6.2% of firms reduced headcount; 18.9% are redesigning roles and 13.9% creating new AI-related jobs. The Accenture read adds a warning: 46% of Singapore companies have not redesigned roles to match the AI they deployed, which means much adoption is still "a technical upgrade rather than an engine for new value". MOM report PDF, Accenture via EDB

How Businesses Are Deploying

The pattern across 2026: enterprises are not buying "AI" as a product. They are standing up centres of excellence, embedding agents into workflows, and attacking sector-specific operational problems.

Centres of Excellence as the anchor move

Source: EDB Q2 2026 AI roundup

Sector deep dives

Banking: DBS is the cited reference case, with a decade-long head start from leadership decisions in 2014: automation embedded in daily workflows, early and heavy investment in data quality, and a culture that rewards experimentation. The Accenture conclusion: AI advantage compounds, and early sustained investment beats fast-following. EDB on DBS

Manufacturing: Sunningdale Tech, a precision plastics maker, deployed AI defect detection with A*STAR's AIMfg centre, cutting manual inspection load. Hyundai's Singapore hub pairs robots with AI on the factory floor. Microsoft and A*STAR signed an MoU on agentic AI for manufacturing. The EDB thesis: manufacturers fail on AI for three reasons (no in-house expertise, poor factory data, uncertainty about live operations), and the ecosystem pairs them with research institutes to cross those barriers. EDB Q2 2026

Connectivity: Singapore Airlines, SATS and Changi Airport Group are exploring AI for air connectivity. Changi T5 (adding ~50 million passengers of capacity) was used by Teo as the flagship example of why "a new terminal alone won't do the job": next-generation air traffic management, baggage flow, runway sequencing. Rolls-Royce signed an MoU for an AI CoE on agentic AI for Power Systems; Thales made Singapore one of three global R&D centres for FlytEDGE. EDB Q2 2026, EDB Q1 2026

Retail and logistics: Sea's LLMs power Shopee features across Southeast Asia at a fraction of commercial LLM cost. Grab is testing robotics in F&B and logistics at Punggol Digital District. EDB on Sea

The Startup Layer

Capital is flowing and the state is building the physical infrastructure for founders. Singapore is the largest AI startup hub in Southeast Asia, and the money is concentrated in infrastructure, agents and applied AI.

Deals worth knowing

Sources: DealStreetAsia on Acrab, EDB on AMI Labs, TechNode × Tracxn, KrASIA.

The homegrown scene

Seedtable tracks 37 funded AI startups in Singapore that have raised US$1.5B between them, mostly Seed to Series A. Names on the leaderboard: Whale (Series C, commerce AI), dConstruct Technologies (Series A), Trax (Series E, retail analytics), Pollo AI (AI video generation), Bifrost AI (synthetic 3D data), Taiger (document intelligence, Series B), CrediLinq AI (embedded finance). Seedtable

Homegrown models matter here too: SEA-LION and MERaLiON are Singapore's open-source, regionally-tuned models, built by the national AI ecosystem. Teo's keynote

Speculation: with Kampong AI opening in 2028 and the NAIIP funnelling SMEs toward adoption, the next two years should produce a visible crop of Singapore-founded vertical AI companies in fintech, healthcare and manufacturing support services.

Money & Policy Levers

Budget 2026 was the largest in Singapore's history, and AI was a named growth driver. If you are building a business here, these are the levers that change your economics.

Sources: Business Times on Budget 2026, SG AI grants directory, Oracle × DISG, Temasek Review 2026 notes.

Governance is a lever too. The Model Governance Framework for Agentic AI, developed with industry and launched at Davos in January 2026, was updated at ATxSummit with case studies from PwC and Workday, plus an OpenClaw case study. PDPC continues to publish AI governance guidance for organisations. Governance frameworks are being updated in months, not years. MDDI, PDPC

Where the Whitespace Is

Read as a business strategist, the data points to specific gaps. The state is paying for adoption; the market is early; the bottlenecks are known and named. Here is where a Singapore business should be looking.

The sharpest read for a data-protection and compliance business

The pattern across all six gaps: every funded adoption push creates a compliance tail. SMEs will adopt AI because the state subsidises it, then discover they need PDPA-compliant data handling, AI governance policies, DPIAs and breach-readiness. The government is simultaneously creating the demand (grants, missions) and the rules (agentic AI framework, PDPC guidance). A business positioned as the trust layer for AI adoption, selling assessment, training and assurance, sits exactly between the two.

Speculation: the winners of the next 24 months will be vendors who package compliance as a prerequisite for grant-funded AI adoption, not as a separate, optional audit. The KPMG move validates the category; the SME scale is the opening.

Sources

All key claims trace to government, EDB, or verified financial press. Figures are as reported by the cited sources as of August 2026.