AI Singapore (AISG) — National Programme for AI Research and Innovation

Institutional Design and National Mandate

AI Singapore (AISG) is the national programme established in May 2017 to anchor Singapore’s artificial intelligence capabilities across the research, innovation, and adoption spectrum. Hosted by the National University of Singapore (NUS) and funded through the National Research Foundation (NRF) with an initial SGD 150 million allocation that has since grown to SGD 500 million cumulatively through NAIS 2.0 supplementary funding, AISG operates as a convening platform that bridges the gap between academic AI research and commercial/government AI deployment—a gap that Singapore’s policymakers identified as the primary bottleneck to national AI competitiveness.

AISG’s organizational model is deliberately hybrid. Rather than building a standalone research institute, AISG operates through a federated network comprising five university AI labs (NUS, NTU, SUTD, SMU, and SIT), six A*STAR research institutes, and over 200 industry partners. This federated structure leverages existing research capabilities while avoiding the institutional overhead of creating a new organization from scratch. AISG’s core team of approximately 250 staff (researchers, engineers, programme managers, and administrative personnel) serves as the connective tissue, managing collaborative projects, providing shared infrastructure, and running national programmes that no single institution could operate independently.

The programme’s executive director reports to a governing council co-chaired by the NRF and the Smart Nation and Digital Government Group (SNDGG), ensuring alignment between AISG’s activities and the government’s broader AI strategy. Funding flows through three channels: core programme funding from NRF (covering AISG operations, shared infrastructure, and national programmes), project-specific grants from participating agencies (funding individual AI projects with defined deliverables), and industry co-funding (where commercial partners contribute matching funds for collaborative R&D projects). The co-funding model ensures that AISG’s work remains commercially relevant—projects that cannot attract industry co-funding are deprioritized regardless of their technical novelty.

The 100 Experiments Programme

AISG’s most visible and commercially impactful initiative is the 100 Experiments (100E) programme, which pairs AI researchers with companies to solve real business problems using AI. The programme provides end-to-end support: problem scoping workshops to define the AI use case, access to AISG’s engineering team for model development, compute resources through the National Supercomputing Centre, and deployment support to operationalize successful solutions. Each experiment receives up to 9 months of AISG engineering support and SGD 250,000 in in-kind resources, with the industry partner contributing data, domain expertise, and deployment infrastructure.

Since its launch in 2018, the 100E programme has completed 280 projects across 18 industry sectors—significantly exceeding its original numerical target. The programme’s success rate (defined as projects that deploy a working AI solution with measurable business impact) is 62%, which AISG benchmarks favorably against the industry average of 35–45% for enterprise AI projects. The higher success rate is attributed to the programme’s rigorous problem scoping process—approximately 40% of proposed projects are refined or redirected during the initial scoping phase, before significant resources are committed.

Sector distribution of 100E projects reflects Singapore’s economic structure: financial services (48 projects), manufacturing (42), healthcare (38), logistics (32), retail (28), and other sectors (92). Notable success stories include the development of an AI-powered quality inspection system for a semiconductor manufacturer that reduced defect detection time by 78% and improved detection accuracy by 15%, an NLP-based customer service automation system for a major bank that handles 35% of customer inquiries without human intervention, and a predictive maintenance system for a ship management company that reduced unplanned vessel downtime by 40%.

The programme’s impact extends beyond individual project outcomes. The 280 completed experiments have created a corpus of applied AI knowledge specific to Singapore’s industrial context—insights about data quality challenges in local manufacturing environments, regulatory constraints on AI deployment in Singapore’s financial sector, and practical integration patterns for AI systems in existing enterprise architectures. This knowledge, documented in AISG’s internal case library and selectively published in academic venues, constitutes an institutional asset that accelerates subsequent projects.

SEA-LION: Southeast Asia’s Foundation Model

AISG’s most technically ambitious project is SEA-LION (Southeast Asian Languages In One Network), a family of large language models trained specifically for Southeast Asian languages and cultural contexts. The project, announced in November 2023 and achieving its first major milestone with the release of SEA-LION v2 in March 2025, addresses a critical gap in the foundation model landscape: the severe underrepresentation of Southeast Asian languages in global LLMs trained primarily on English, Chinese, and European language data.

SEA-LION v2 is available in three sizes: 3B parameters (suitable for edge deployment and resource-constrained applications), 8B parameters (balanced capability and efficiency), and 70B parameters (maximum capability for complex reasoning and generation tasks). The model was trained on a custom-curated dataset of 2.8 trillion tokens comprising text in 11 Southeast Asian languages (Malay/Indonesian, Thai, Vietnamese, Tagalog, Burmese, Khmer, Lao, Javanese, Sundanese, Tamil, and Mandarin as used in Singapore/Malaysia) plus English. The training dataset was assembled through a combination of web crawling (with language-specific quality filtering), partnerships with national libraries and archives across ASEAN, and commissioned data collection targeting underrepresented languages with limited digital text corpora.

The model’s performance on Southeast Asian language benchmarks significantly exceeds that of global foundation models. On the SEA-Bench evaluation suite (developed by AISG specifically for multilingual Southeast Asian assessment), SEA-LION 70B achieves an average score of 78.4 across all 11 languages, compared to 62.1 for GPT-4 and 58.7 for Claude 3 on the same benchmark. The performance advantage is most pronounced in lower-resource languages: for Burmese, SEA-LION outperforms GPT-4 by 31 percentage points; for Khmer, by 28 points; and for Lao, by 35 points. For higher-resource languages (Malay, Thai, Vietnamese), the advantage is smaller but still significant at 8–15 percentage points.

SEA-LION is released under an open-source license (Apache 2.0) and is hosted on Hugging Face for community access. The open-source release strategy reflects AISG’s assessment that the commercial value of foundation models increasingly lies in fine-tuning and application layer rather than in the base model itself—making open release a strategy that maximizes ecosystem value while creating opportunities for Singapore-based companies to build commercial applications on top of the open model. As of Q1 2026, SEA-LION has been downloaded 450,000 times and is used in production by 85 organizations across 8 ASEAN countries.

AI Apprenticeship and Talent Development

AISG’s AI Apprenticeship Programme (AIAP) is Singapore’s most structured pathway for developing applied AI professionals. The 9-month full-time programme accepts cohorts of 40–50 apprentices from diverse educational and professional backgrounds—the only requirements are basic programming proficiency and strong analytical aptitude. Apprentices receive a monthly stipend of SGD 3,500 while working on real AI projects under the mentorship of AISG’s senior AI engineers.

The programme’s curriculum is unusual in its emphasis on full-stack AI engineering rather than pure data science. Apprentices learn not just model development (which occupies approximately 30% of the curriculum) but also data engineering (25%), MLOps (20%), and product/project management for AI (25%). This full-stack approach reflects the market reality that most AI roles require capabilities beyond model training—data pipelines must be built, models must be deployed and monitored, and AI projects must be managed within organizational contexts.

Since its inception in 2018, AIAP has graduated 380 apprentices across 12 cohorts. Employment outcomes are strong: 92% of graduates secure AI-related positions within 3 months of programme completion, with a median starting salary of SGD 6,500 monthly. Employer feedback, collected through annual surveys, rates AIAP graduates as “production-ready” from day one—a distinction from typical university graduates who require 6–12 months of workplace-based skill development. The programme’s alumni network, spanning 180 organizations, has become an informal talent marketplace where AISG alumni recommend positions and provide peer support.

AISG also operates the AI for Everyone (AI4E) and AI for Industry (AI4I) programmes, which provide shorter-duration AI literacy and skills training to broader audiences. AI4E, a free online course covering AI fundamentals and societal implications, has been completed by 82,000 participants since 2019. AI4I, a 7-week part-time course providing hands-on AI development skills, has trained 8,500 working professionals. Together with AIAP, these programmes constitute Singapore’s primary pipeline for AI human capital development—a pipeline that NAIS 2.0 targets to expand by 40% through increased programme capacity and new specialized tracks in generative AI, AI safety, and AI governance.

Research Commercialization and Technology Transfer

AISG’s research portfolio spans fundamental AI research (conducted primarily through the university network) and applied research (conducted through collaborative projects with industry and government partners). The programme supports approximately 150 active research projects at any given time, with annual research expenditure of SGD 65 million. Research priorities align with NAIS 2.0’s strategic focus areas: trustworthy AI (including fairness, robustness, and explainability), efficient AI (including model compression, few-shot learning, and energy-efficient training), and multimodal AI (including vision-language models, speech processing, and cross-modal retrieval).

Technology transfer from research to deployment operates through three mechanisms. Direct commercialization occurs when AISG research produces intellectual property that is licensed to industry partners or spun out as startups—six AISG spinoffs have been incorporated since 2019, with two achieving Series A funding. Embedded deployment places AISG researchers within industry partner organizations for 6–12 months to transfer specific AI capabilities and build internal team capacity. Open-source release makes AISG research outputs freely available for community use—AISG has published 45 open-source AI tools and libraries on GitHub, with the most widely used being the AI Verify testing toolkit (co-developed with IMDA) and the SEA-LION model family.

The technology transfer success rate—measured as the proportion of research projects that produce deployable AI capabilities adopted by at least one organization—is 34%, which AISG benchmarks against the global average of 20–25% for government-funded AI research programmes. The higher success rate reflects AISG’s programme design, which embeds commercial relevance criteria into research project selection and maintains ongoing industry engagement throughout the research lifecycle rather than treating commercialization as an afterthought.

AISG’s federated model, combining national coordination with distributed execution across universities and research institutes, has proven effective for a small nation seeking AI competitiveness without the resources for a standalone national AI lab on the scale of the UK’s DeepMind or the U.S.’s national labs. The model’s scalability is being tested as NAIS 2.0 expands AISG’s mandate and budget—whether the federated approach can maintain its agility and commercial relevance at twice its current scale is an open question that will significantly influence Singapore’s AI trajectory through the end of the decade.

Extended Analysis and Contextual Intelligence

The extended analysis of this domain draws on Singapore’s unique position as a small, open, highly developed economy that consistently punches above its weight in technology, governance, and institutional innovation. The city-state’s approach to national development—combining strategic vision with pragmatic execution, sustained investment with rigorous evaluation, and international engagement with domestic capability building—provides the institutional foundation for the programmes and policies examined in this analysis.

Singapore’s governance model, characterized by strong institutional capacity, meritocratic talent management, and evidence-based policy development, creates conditions that are difficult to replicate in other jurisdictions but that provide instructive lessons for governments and organizations worldwide. The model’s emphasis on long-term planning, institutional learning, and adaptive management has produced outcomes that consistently exceed what Singapore’s resource base and population size would predict, establishing the city-state as a reference case for effective governance in the digital age.

The economic context shapes both the opportunities and constraints for development in this domain. Singapore’s GDP per capita of approximately SGD 85,000 provides the fiscal resources for public investment while creating a high-cost operating environment that demands productivity and innovation. The economy’s openness to trade, investment, and talent creates opportunities for international collaboration while exposing domestic industries to global competitive pressures. The demographic profile—an aging population, a diverse multicultural society, and significant reliance on international talent—creates both challenges and opportunities for workforce development and social policy.

Technology evolution continues to reshape the possibilities for institutional performance and service delivery. Artificial intelligence, cloud computing, distributed ledger technology, and the Internet of Things are collectively transforming how governments operate, how businesses compete, and how citizens interact with institutions. Singapore’s approach of being an early but disciplined adopter of technology—investing in understanding before committing to deployment, and evaluating outcomes rigorously once deployed—provides a model for technology governance that balances innovation with risk management.

The international dimension remains central to Singapore’s strategy in this domain. As a small nation dependent on global connectivity for economic prosperity and security, Singapore cannot afford to operate in isolation. International partnerships, regulatory cooperation, standard-setting participation, and knowledge exchange all contribute to the city-state’s ability to maintain capabilities that exceed what domestic resources alone could sustain. The diplomacy of technology cooperation—building relationships through shared standards, mutual recognition, and collaborative research—has become a significant dimension of Singapore’s international engagement strategy.

Looking toward the remainder of the Smart Nation 2.0 implementation period and beyond, the analysis identifies several themes that will shape development in this domain. The integration of AI capabilities into routine institutional operations will continue to deepen, creating both efficiency gains and governance challenges. The sustainability imperative will increasingly influence investment decisions, technology choices, and performance measurement. The regional dimension will grow in importance as ASEAN integration deepens and cross-border digital flows increase. And the talent challenge will remain the binding constraint that ultimately determines the pace and scope of achievement.

The intelligence presented in this analysis is designed to support decision-makers who need to understand Singapore’s trajectory in this domain—whether for investment decisions, policy analysis, competitive assessment, or academic research. The Vanderbilt Terminal’s commitment to data-dense, authoritative intelligence ensures that this analysis provides the factual foundation and analytical framework needed for informed judgment, while acknowledging the uncertainties and alternative interpretations that honest intelligence assessment requires.

Supplementary Intelligence and Data Context

Additional data context supports the analytical assessments presented throughout this intelligence product. Singapore’s statistical infrastructure, maintained by the Department of Statistics and supplemented by sector-specific data collection by statutory boards and regulatory agencies, provides the quantitative foundation for performance monitoring and international comparison. The data quality framework governing government statistical production ensures that reported metrics meet international standards for accuracy, timeliness, and methodological rigor.

The policy context for this analysis is provided by Singapore’s national planning framework, which operates on multiple time horizons. The annual budget cycle allocates resources and sets near-term priorities. The five-year Research, Innovation and Enterprise plan provides the medium-term investment framework for technology and innovation. The Long-Term Plan Review, conducted at approximately ten-year intervals, establishes the strategic direction for infrastructure and land use planning across 30-50 year horizons. This multi-horizon planning approach ensures that near-term decisions are informed by long-term strategic considerations.

The institutional context reflects Singapore’s distinctive governance model, which combines strong state capacity with pragmatic engagement with market forces. Government agencies operate with clear mandates, measurable objectives, and accountability mechanisms that drive performance. Market participation is encouraged through regulatory frameworks that establish clear rules while maintaining space for innovation and competition. The interaction between state direction and market dynamism produces outcomes that neither pure state planning nor pure market allocation could achieve independently.

Industry perspectives complement government data in providing comprehensive intelligence. Corporate reporting by Singapore-listed companies provides financial and operational data that reveals market dynamics from the private sector perspective. Industry association publications aggregate member data to provide sector-level intelligence. Professional services firms contribute analytical frameworks and benchmark data from their advisory engagements. Media coverage provides real-time intelligence on developments that may not yet be reflected in official statistics or corporate reporting.

The geopolitical context increasingly shapes the development trajectory for this domain. Singapore’s position as a small, trade-dependent city-state at the intersection of major power interests creates both opportunities and risks. The opportunity lies in Singapore’s ability to serve as a neutral platform for international cooperation and commercial exchange. The risk lies in the potential for great power competition to fragment the international systems on which Singapore’s prosperity depends. Managing this geopolitical positioning requires the diplomatic skill and strategic clarity that have characterized Singapore’s international engagement since independence.

The demographic context adds another analytical dimension. Singapore’s resident population of 5.9 million is aging rapidly, with the proportion of residents aged 65 and above projected to increase from 19% in 2025 to 25% by 2030. This demographic shift affects demand patterns for government services, workforce composition and availability, healthcare system requirements, and social security system sustainability. The integration of technology solutions to address aging-related challenges is a cross-cutting theme that connects Smart Nation programmes with social policy objectives.