Document Code: SG-E-49 Full Title: The Singapore Startup Ecosystem — From Block 71 to the AI Era: State-Directed Entrepreneurship, Unicorn Creation, and the Limits of Managed Innovation (2010–2026) Coverage Period: 2010–2026 Level Designation: Level 2 Status: [COMPLETE] Primary Sources Consulted:

NUS Enterprise, Block 71 — Ayer Rajah Crescent Innovation Precinct programme documentation, 2011–2026; NUS Enterprise Annual Reports and impact assessments
Enterprise Singapore (formerly SPRING Singapore and IE Singapore), Startup SG programme documentation: Startup SG Founder, Startup SG Tech, Startup SG Equity, Startup SG Talent, 2017–2026; Enterprise Singapore Annual Reports 2018–2026
Ministry of Trade and Industry (MTI), Singapore Budget Statements (relevant sections on SME and startup support), 2010–2026; MTI committee of supply speeches on entrepreneurship
JTC Corporation, Ayer Rajah precinct development materials; JTC innovation district planning documentation, 2010–2020
SingTel Innov8, corporate venture capital portfolio and programme materials, 2011–2020; SingTel Group Annual Reports (venture capital sections), 2011–2020
Vertex Ventures, Fund I–V public documentation; Vertex Holdings (Temasek subsidiary) annual portfolio reviews, 2014–2026
Heliconia Capital Management (subsidiary of Temasek), portfolio and programme documentation, 2012–2026
Economic Development Board (EDB), EDBI portfolio documentation; EDB investment facilitation for startup ecosystem, 2010–2026; EDB media releases on startup investment commitments
Grab Holdings, IPO Prospectus (NASDAQ listing via Altimeter Growth Corp SPAC merger, December 2021); Grab Annual Reports 2022–2025
Sea Limited, NYSE IPO Prospectus (2017) and subsequent Annual Reports 2018–2025; Sea Limited investor-day presentations
Carro Automotive, funding-round press releases and company materials, 2015–2026; media coverage of Carro's Series C (US$360M, June 2021, SoftBank Vision Fund 2 lead, valuation just above US$1 billion) and pre-IPO rounds (US$100M sought at >US$1.5B valuation, April 2024)
Carousell Group, funding-round announcements and company materials, 2012–2026; Carousell media releases on STIC Investments-led September 2021 round (US$100M at US$1.1B valuation) and merger with 701Search/Telenor (November 2019)
Monetary Authority of Singapore (MAS), Digital Payment Token (DPT) service provider licensing documentation; MAS media releases and enforcement actions relating to Three Arrows Capital (3AC), FTX/FTX Singapore (Quoine Pte Ltd), and Terra-Luna ecosystem, 2022–2023
MAS, Review of the Digital Payment Token (DPT) Services Regulatory Framework, 2022; MAS Enforcement Actions register, 2022–2024
Jungle Ventures, Sequoia Southeast Asia, B Capital Group fund documentation and portfolio press releases, 2015–2026 (local VC ecosystem)
NRF (National Research Foundation), National Startup Ecosystem Reports, various years; NRF, Deep Tech Translation Programme documentation, 2020–2026
e27 and DealStreetAsia, Southeast Asia Startup Ecosystem Reports, 2015–2026 (trade media; corroborate with primary sources)
Startup Genome, Global Startup Ecosystem Report, 2022, 2023, 2024, 2025 — Singapore rankings and ecosystem benchmarks
The Straits Times, Business Times, CNA, and The Edge Singapore contemporaneous reporting on Singapore unicorns, crypto fallout, and AI startup wave, 2010–2026
Philip Yeo (as told to Peh Shing Huei), Neither Civil Nor Servant (Singapore: Straits Times Press, 2018) — on A*STAR commercialisation philosophy
Israel Innovation Authority, State of Israeli Tech reports 2022–2024; NASSCOM, India Tech Ecosystem Report 2023; comparative benchmarks for Stockholm (Vinnova reports) and Shanghai (Zhangjiang Hi-Tech Park data)
Singapore Economic Development Board / NUS Enterprise / Enterprise Singapore, ecosystem-overview materials, 2024–2026; Startup Genome, Global Startup Ecosystem Report 2024 (Singapore ranked #7 globally; US$144 billion ecosystem value Jul 2021 – Dec 2023)

Related Documents:

SG-E-01: The Economic Development Board
SG-E-15: Research, Innovation, and Enterprise
SG-E-16: A*STAR
SG-E-25: The Digital Economy
SG-E-27: Committee on the Future Economy
SG-E-36: Crypto, Fintech, and the Family Office Hub
SG-E-46: The Industrial Strategy — From Goh Keng Swee to Tan See Leng
SG-E-47: Wage Models — IWI, PWM, and WCS
SG-O-01: The AI Mega Trend — Singapore's Strategy, Stakes, and Vulnerabilities
SG-O-12: AI Governance Deep-Dive
SG-O-14: Jobs Versus AI in Singapore — The Labour-Market Reckoning
SG-D-17: Technology, Innovation, and the Smart Nation
SG-B-09: Lawrence Wong Transition
SG-M-09: The Developmental State — Singapore's Variant

Version Date: 2026-05-16

1. Key Takeaways

Block 71 — the Ayer Rajah Crescent innovation precinct launched in 2011 — is the single most legible symbol of Singapore's deliberate, state-mediated approach to startup ecosystem construction. What became a cluster of repurposed JTC industrial buildings housing over 1,000 startups, 50 accelerators, and multiple corporate innovation labs did not emerge organically from a critical mass of risk-tolerant entrepreneurs. It was designed, seeded, and continuously scaffolded by NUS Enterprise, JTC Corporation, the Economic Development Board, and a succession of government programmes. The Block 71 model — subsidised space, curated tenants, government-funded mentorship, and co-investment vehicles — became Singapore's answer to the question of whether a city-state without a natural technology hinterland could manufacture an innovation economy. By 2020 the answer was, at minimum, "yes, but at a price."
The Startup SG programme, launched formally in 2017, consolidated what had been a fragmented set of agency-specific startup support schemes into a single, branded framework with four pillars: Startup SG Founder (co-funding grants for first-time entrepreneurs), Startup SG Tech (deep-tech commercialisation), Startup SG Equity (government co-investment via SEEDS and i.JAM programmes), and Startup SG Talent (employment-pass facilitation for foreign founders). Enterprise Singapore, formed from the 2018 merger of SPRING Singapore and IE Singapore, became the administrative hub of this ecosystem. The programme was calibrated to correct three market failures that Singapore's planners identified: under-investment in high-risk early-stage ventures, under-representation of Singaporean founders relative to foreign talent, and insufficient translation of A*STAR and university research into commercially viable products.
Singapore produced eight unicorns (privately held companies valued at US$1 billion or more) between 2014 and 2022, a record that, on a per-capita basis, placed it among the world's leading startup ecosystems. The six most prominent — Grab, Sea (formerly Garena), Carro, Carousell, Razer, and NinjaVan — collectively represented a Southeast Asian internet economy that Singapore both incubated and connected to global capital markets. Yet the unicorn roster also revealed a structural asymmetry: most of Singapore's largest technology companies were built on Southeast Asian consumer markets, not on Singapore's domestic market. Singapore was the headquarters, the regulatory anchor, and the capital-raising node; the actual business was elsewhere. This divergence between where value was created and where it was captured has become one of the ecosystem's most debated structural features.
The 2022–2024 crypto crisis — the collapse of Three Arrows Capital (founded by Singapore-based Su Zhu and Kyle Davies), the Terra-Luna implosion, and the FTX Singapore fallout — was the most significant reputational shock to Singapore's financial and technology credibility since the 2008–2009 Global Financial Crisis. Three Arrows Capital had been a Singapore-registered fund; Terra's co-founder Do Kwon had maintained Singapore connections; FTX Singapore (trading as Quoine Pte Ltd) held a Major Payment Institution licence from MAS. The combined losses across these entities ran to tens of billions of dollars globally — Terra-Luna alone destroyed roughly US$40–60 billion in market value in May 2022, and 3AC's bankruptcy filing in July 2022 listed approximately US$3.5 billion in creditor claims (Singapore-specific retail-investor losses across the three episodes have never been comprehensively tabulated by MAS). MAS's response — licensing tightening, fit-and-proper enforcement, and public statements distinguishing "legitimate digital-asset innovation" from speculative excess — reshaped the crypto-fintech ecosystem and accelerated the pivot of government policy attention toward AI-native startups.
The 2024–2026 AI startup wave represented a qualitative shift in the ecosystem's composition. Earlier waves had been dominated by consumer-internet, marketplace, and fintech models. The AI wave was characterised by foundation-model fine-tuning ventures, enterprise AI deployment companies, AI-native vertical SaaS, and deep-tech spinouts from A*STAR's Institute for Infocomm Research (I²R) and the National University of Singapore. The Government's SGD $1 billion AI compute fund (launched 2023), the National AI Strategy 2.0, and the AI Singapore (AISG) 100E (100 Experiments) and AIAP (AI Apprenticeship Programme) created a pipeline of AI-ready talent and proof-of-concept deployments. By 2026, AI-adjacent startups constituted a growing share of new incorporations at Block 71 — NUS Enterprise and Microsoft jointly announced in October 2025 a programme to fast-track 150 AI startups through the Startup SG Tech grant over three years, an indicator of the cohort's prominence .
Singapore's sovereign-venture architecture — the network of state-linked investment vehicles including Vertex Ventures (Temasek subsidiary), Heliconia Capital (Temasek), EDBI (EDB investment arm), and Pavilion Capital — distinguishes it from almost every other startup ecosystem. These vehicles do not simply co-invest; they set market norms, signal investible sectors, and provide patient capital that commercial VCs, subject to fund-cycle pressures, cannot always match. The architecture has been credited with sustaining the ecosystem through periods when global VC sentiment toward Southeast Asia cooled (notably 2022–2023). It has also been criticised for crowding out independent fund managers, creating pricing distortions, and sustaining companies that commercial discipline would have forced to pivot or close.
Comparative analysis places Singapore in a distinct category among global startup ecosystems: high on capital access, regulatory quality, and international connectivity; constrained on domestic market size, risk-appetite culture, and tolerance for founder failure. Tel Aviv, the most frequently cited comparator, generates disproportionate deep-tech output through military technology transfer (Unit 8200 alumni, RAFAEL spin-offs) and a cultural comfort with failure that decades of existential national experience has forged. Bangalore generates volume through a vast engineering talent pool and domestic market scale. Stockholm generates per-capita unicorn density through a combination of flat organisational culture and early gaming/music digitisation. Singapore has sought to approximate elements of each model, but the underlying cultural and structural conditions — risk aversion reinforced by housing costs, a meritocracy ideology that penalises failure, and a labour market that rewards MNC employment over entrepreneurial risk — remain durable constraints.

2. The Record in Brief

Singapore's startup ecosystem as it exists in 2026 is approximately fifteen years old as a coherent, state-supported construct. It has antecedents — the Economic Development Board's early-stage investment programmes, A*STAR's commercialisation offices, the National University of Singapore's entrepreneurship clubs, and a handful of early internet companies from the late 1990s — but the transformation of Singapore from a predominantly MNC-and-GLC economy into one in which startups constituted a visible and increasingly significant economic constituency dates from the 2010–2011 period.

The proximate catalyst was the convergence of three developments. First, the Economic Strategies Committee (ESC) report of February 2010, chaired by Finance Minister Tharman Shanmugaratnam, argued for an "innovation-driven economy" and identified entrepreneurship as a structural gap. Second, the National Research Foundation's (NRF) Early-Stage Venture Fund (ESVF) and other pre-competitive funding mechanisms were scaled up. Third, a group of NUS administrators, JTC planners, and EDB officers converged on a specific site — a cluster of one-storey flatted factory buildings at 71 Ayer Rajah Crescent — and decided to turn it into an innovation precinct rather than lease it to industrial tenants.

The Block 71 opening in 2011 was, in retrospect, a founding moment for a new phase of Singapore's economic policy. Over the following decade, the ecosystem grew through successive waves: the consumer-internet and e-commerce wave (2011–2016), the fintech wave (2016–2020), the deep-tech and biotech wave (2018–2022, overlapping with fintech), and the AI-native wave (2024–present). Each wave was partly organic — reflecting global technology trends — and partly managed, as government programmes recalibrated their priorities and funding allocations to track the technology frontier.

The governance architecture evolved accordingly. SPRING Singapore and IE Singapore, the two agencies historically responsible for SME development and internationalisation, were merged into Enterprise Singapore in 2018, providing a single institutional home for the Startup SG framework. The Monetary Authority of Singapore simultaneously developed its regulatory sandbox and digital payment token licensing regime, providing the legal infrastructure for the fintech wave. NRF's deep-tech translation programmes were consolidated and expanded. The result was an unusually coherent — though not always internally consistent — institutional ecosystem in which multiple agencies played defined roles with limited duplication.

The 2022–2024 crypto crisis disrupted this trajectory. The failures of Three Arrows Capital, Terra-Luna, and FTX Singapore were not merely commercial failures; they were governance failures that raised questions about MAS's supervisory effectiveness and Singapore's due-diligence culture. The government's response — tighter DPT licensing, stricter fit-and-proper standards, and a deliberate policy signal that Singapore would support "responsible" digital-asset innovation rather than speculative excess — reoriented the ecosystem's regulatory posture. By 2025 the fintech and crypto sub-ecosystems had contracted significantly in headcount and capital, while AI and deep-tech had expanded.

By 2026, Singapore's startup ecosystem had attracted approximately 4,500 startups (across varying stages), more than 150 active venture capital funds, and a pipeline of talent fed by NUS, NTU, SMU, and SUTD entrepreneurship programmes. The eight unicorns produced between 2014 and 2022 remained the ecosystem's most visible achievements. Block 71 had expanded to include satellite precincts at JTC LaunchPad @ one-north and Punggol Digital District. And the government's investment in AI compute infrastructure — part of the National AI Strategy 2.0 launched in December 2023 — was beginning to attract a new generation of AI-native founders who had previously built or trained elsewhere.

3. Timeline 2010–2026

2010 — Economic Strategies Committee report recommends innovation-driven economy. NRF scales Early-Stage Venture Fund (ESVF). EDB and NUS Enterprise begin planning for Ayer Rajah Crescent repurposing.

2011 — Block 71 opens at Ayer Rajah Crescent, Queenstown. NUS Enterprise, JTC Corporation, and SingTel Innov8 are founding partners. Initial tenants number approximately 40 startups. Garena (later Sea Limited), already incorporated in 2009, begins regional expansion from Singapore.

2012 — Carousell founded by NUS students Marcus Tan, Lucas Ngoo, and Quek Siu Rong (NUS Overseas Colleges programme). First cohort of NUS Overseas Colleges (NOC) Silicon Valley alumni return. SingTel Innov8 invests in first portfolio companies.

2013 — Block 71 tenants exceed 250 startups. Grab (then MyTeksi) relocates headquarters from Malaysia to Singapore. NRF's Technology Incubation Scheme (TIS) expanded. Wired magazine names Block 71 "the world's most tightly packed entrepreneurial ecosystem."

2014 — Grab raises US$10 million Series A led by Tiger Global. Sea Limited (Garena) raises major round; gaming revenues fund e-commerce pivot. Government launches the Committee on the Future Economy preparatory consultations.

2015 — Carro founded. NinjaVan founded. ShopBack founded. The Singapore FinTech Festival (SFF) concept being developed. MAS introduces regulatory sandbox framework. Block 71 Singapore overseas editions announced: Block 71 Jakarta (2015) and Block 71 San Francisco (2016).

2016 — Committee on the Future Economy convened. MAS launches the Regulatory Sandbox for fintech. First Singapore FinTech Festival (SFF) held in November 2016, attracting 13,000 attendees. Grab launches GrabPay.

2017 — CFE reports in February, endorsing startup ecosystem as growth vector. Startup SG programme formally launched, consolidating existing schemes under Enterprise Singapore (in formation). Sea Limited IPO on NYSE on 20 October 2017 at US$15 per American Depositary Share, raising approximately US$884 million — Singapore's first major technology IPO on a US exchange.

2018 — Enterprise Singapore formed from merger of SPRING Singapore and IE Singapore. Startup SG Talent, Founder, Tech, and Equity pillars operationalised. Grab's Series G round closed in early 2018 jointly led by SoftBank and Didi at a pre-round valuation of approximately US$6 billion; in June 2018 Toyota then led a US$1 billion investment that valued Grab at over US$10 billion. Block 71 expansion to Sandbox @ Block 71.

2019 — Razer achieves unicorn status (HKEX IPO 2017; unicorn designation precedes listing). Grab secures US$1.46 billion from the SoftBank Vision Fund in March 2019, bringing its Series H total to over US$4.5 billion and valuing Grab at approximately US$14 billion. Sea Limited's Shopee achieves market leadership in multiple Southeast Asian countries. MAS issues guidelines on digital token offerings.

2020 — COVID-19 accelerates digital adoption. Grab and Gojek merger discussions (ultimately unsuccessful). Sea Limited's market capitalisation surges past US$90 billion by late 2020 (stock up 385% in 2020) and continues climbing into 2021, making it briefly the most valuable Southeast Asian company. MAS grants four digital bank licences.

2021 — Sea Limited's stock peaks at US$366.99 on 19 October 2021, taking its market capitalisation close to US$200 billion. Grab merges with Altimeter Growth Corp (SPAC) and lists on NASDAQ on 2 December 2021, valued at approximately US$40 billion at listing — the largest SPAC merger on record at the time. Carro raises US$360 million Series C in June 2021, led by SoftBank Vision Fund 2, reaching unicorn status at just over US$1 billion. Carousell raises US$100 million in September 2021 led by STIC Investments at a US$1.1 billion valuation, formally entering the unicorn club. NinjaVan raises a US$578 million Series E (Alibaba-backed) in September 2021 at a >US$1 billion valuation. PatSnap raises a US$300 million Series E (SoftBank Vision Fund 2 / Tencent) in March 2021, reaching unicorn status. Advance Intelligence Group raises a US$400M+ Series D in September 2021 (SoftBank Vision Fund 2 / Warburg Pincus) at a >US$2 billion valuation. Three Arrows Capital at peak AUM reportedly manages approximately US$10 billion.

2022 — May 2022: Terra-Luna collapse. Approximately US$40–60 billion in aggregate market value destroyed globally within weeks (LUNA token alone fell from a >US$40 billion peak market cap to under US$500 million in days). Do Kwon, who had run Terraform Labs from Singapore, flees the jurisdiction. 27 June 2022: a British Virgin Islands court orders Three Arrows Capital (BVI-incorporated but Singapore-headquartered) into liquidation; Su Zhu and Kyle Davies leave Singapore. 30 June 2022: MAS reprimands Three Arrows Capital Pte. Ltd. for providing false information and exceeding the S$250 million AUM threshold for a Registered Fund Management Company between July–September 2020 and November 2020–August 2021. November 2022: FTX global collapse; FTX's Singapore-registered subsidiary Quoine Pte Ltd (operator of the Liquid exchange) was at the time exempt from licensing while its DPT licence application was under MAS review, not a Major Payment Institution licensee. MAS issues public statements clarifying that FTX.com was never licensed in Singapore.

2023 — MAS tightens DPT licensing; multiple crypto entities exit Singapore or surrender licences. National AI Strategy 2.0 launched 4 December 2023; Budget 2024 subsequently confirmed an investment of more than S$1 billion over five years to back NAIS 2.0, including up to S$500 million dedicated to securing high-performance compute for AI innovation. AI Singapore's 100 Experiments programme passes 100+ deployments since 2017 (with 700+ companies engaged across all AISG outreach). Block 71 AI cluster begins forming. Grab full-year 2023 revenue grew 65% year-on-year to US$2.36 billion. March 2023: Do Kwon arrested at Podgorica airport, Montenegro, with a falsified Costa Rican passport.

2024 — Singapore ranks #7 globally in the Startup Genome Global Startup Ecosystem Report 2024 (up one place from 2023), the leading ecosystem in Asia ahead of Beijing, Seoul, Tokyo and Shanghai, with US$144 billion in ecosystem value created July 2021 – December 2023. AI-native startups constitute growing share of new incorporations. Lawrence Wong PM era begins; Forward Singapore economic review incorporates startup ecosystem sustainability. Several A*STAR spinouts achieve early-stage funding. September 2024: Enterprise Singapore and Google Cloud announce a joint AI accelerator (Google for Startups Accelerator: AI First Singapore).

2025 — Singapore government high-performance compute allocation under NAIS 2.0 begins disbursements. Multiple AI foundation-model startups incorporated. ShopBack restructures and pivots to AI-powered commerce. October 2025: Microsoft, Enterprise Singapore and NUS Enterprise launch a joint accelerator to fast-track 150 AI startups through the Startup SG Tech grant over three years. Block71 Singapore has been home to over 1,600 ventures cumulatively (NUS Enterprise figure).

2026 — SGD $1 billion AI compute fund deployment ongoing. AI startup wave continues; new cohort of deep-tech AI companies completing Series A rounds. MAS crypto licensing framework fully operational; DPT sector stabilised at smaller scale. NUS Enterprise and NTU entrepreneurship programmes feeding AI-trained cohort into ecosystem.

4. The Pre-2010 Architecture — A*STAR, TIF, EDB Entrepreneurship Tracks

Understanding why Block 71 was necessary requires understanding what came before — and why it was insufficient. Singapore had been investing in research and commercialisation infrastructure for over two decades before 2010, but the results in terms of commercially successful startups had been disappointing relative to the investment.

The Agency for Science, Technology and Research (ASTAR), established in 2002 from the National Science and Technology Board (NSTB) and the Institute of Technical Education (ITE) precursor structures, was Singapore's primary instrument for building a research ecosystem that could produce commercially translatable outputs. By 2010, ASTAR hosted twenty research institutes covering biomedical sciences, computing, advanced manufacturing, and engineering. Its annual research expenditure exceeded SGD $1.5 billion. Yet the translation of ASTAR research into Singapore-domiciled startups remained thin. The institute structure, optimised for publication metrics and technology licensing to MNCs, did not naturally produce the kind of founder-centric, commercially oriented research that typical VC-backed startups require. As Philip Yeo, the founding chairman of ASTAR, later acknowledged in Neither Civil Nor Servant (2018), the commercialisation culture was a persistent challenge — one that institutional design alone could not resolve.

The Technology Incubation Scheme (TIS), managed by NRF from 2008, provided co-investment with qualified incubators at a 85:15 (government:incubator) ratio for early-stage deep-tech startups. The Business Angel Scheme (BAS) and SPRING SEEDS programme provided additional early-stage co-investment, matching private angel investments in qualifying startups. These were sound instruments but operated at modest scale and lacked the precinct-cluster dynamic that could create network effects.

The EDB's entrepreneurship tracks — including the Technopreneurship 21 (T21) programme launched in 1999 and various high-tech venture facilitation efforts — reflected the EDB's characteristic strategy of importing foreign expertise to seed local capability. T21 attracted several American venture capital firms to Singapore and facilitated some early-stage investment activity, but the bubble burst in 2001 before the ecosystem could achieve critical mass, and many of the foreign VC presences retreated.

NUS Enterprise, established formally as an office within the National University of Singapore, launched the NUS Overseas Colleges (NOC) programme in 2002 — a year-long entrepreneurship immersion in which NUS students were embedded at Silicon Valley, Shanghai, Stockholm, and other startup hubs while interning at high-growth companies. NOC was quietly transformative. Over the following decade it produced a cohort of Singaporean graduates who had experienced Silicon Valley's cultural norms first-hand, had built networks with foreign founders and investors, and returned to Singapore with a materially different orientation toward risk and entrepreneurship. Carousell's founding team were NOC Silicon Valley alumni. Numerous other Block 71 founding-era entrepreneurs had NOC exposure.

The Singapore Infocomm Technology Federation (now SGTech), iJAM (Incubator for Joyful Arts and Media), and various Media Development Authority (now IMDA) programmes served niche ecosystem roles. But the overall pre-2010 architecture was fragmented, under-resourced relative to its ambitions, and lacked a physical concentration — a place — that could create the serendipitous collisions central to innovation ecosystems everywhere from Silicon Valley to Tel Aviv to Zhongguancun.

5. The 2011 Block 71 Founding — NUS Enterprise, SingTel Innov8, JTC

Block 71 was not, initially, a grand plan. It was a repurposing decision. JTC Corporation, Singapore's industrial landlord, managed a cluster of one-storey flatted factory buildings at 71 Ayer Rajah Crescent in the Queenstown planning area, originally built in the 1990s to house light-manufacturing tenants. By 2010 these tenants were migrating to lower-cost locations, and the buildings were underutilised. NUS Enterprise, casting around for a physical site to anchor its startup programming, saw an opportunity.

The founding partnership was tripartite: NUS Enterprise provided programme content, mentorship networks, and student pipelines; JTC provided the physical space at subsidised rates and managed the building infrastructure; SingTel Innov8, the corporate venture capital arm of Singapore Telecommunications (itself majority Temasek-owned), provided seed capital, corporate validation, and connections to SingTel's commercial ecosystem. The combination was deliberate — a university anchor, a state industrial landlord, and a corporate VC — designed to address the three most common failure modes of university technology parks: lack of physical space, lack of commercial capital, and lack of market connections.

The precinct opened in 2011 under the name "Block 71" (the address of the anchor building) and attracted approximately 40 startups in its first year. The physical environment was deliberately unglamorous — exposed piping, industrial flooring, shared amenities — in implicit contrast to the polished corporate campuses of Singapore's MNC tenants. This aesthetic choice was read by participants as a signal: this is a place for building, not for performing.

What made Block 71 work as a catalyst — in ways that previous Singapore innovation parks had not — was the density of accelerator and investor presence alongside the startups. Joyful Frog Digital Incubator (JFDI), one of Singapore's first accelerator programmes, was an anchor tenant. NUS Enterprise's programmes operated from the site. SingTel Innov8's team was physically present. Over subsequent years, other corporate innovation arms — from DBS Bank, Unilever, and international players including Vertex Ventures — co-located or maintained presence at the precinct. The critical mass of dealmakers, mentors, and operators in a single small physical footprint generated the network effects that earlier, more dispersed Singapore innovation infrastructure had failed to produce.

By 2013 Block 71 had expanded into adjacent buildings (73, 75, and 79 Ayer Rajah Crescent), tenants exceeded 250 startups and 30 accelerators, and Wired magazine's profile had given it international recognition. The government capitalised on this visibility by launching overseas Block 71 editions: Block 71 Jakarta in 2015 (in partnership with Cyber Creative Institute), Block 71 San Francisco in 2016, Block 71 New York in 2018, and Block 71 Ho Chi Minh City in 2018. These overseas nodes functioned as bridgeheads for Singapore-based startups entering new markets and for channelling foreign startup talent toward Singapore.

The Block 71 model's political economy deserves scrutiny. The precinct was deeply subsidised: below-market rents, government-funded programming, and state-linked co-investment. This made it accessible to early-stage founders who could not afford Singapore's commercial real estate market, but it also meant that the ecosystem's viability depended on continued government commitment. When JTC rents were periodically marked to market or space was reallocated to competing uses, founders complained. The ecosystem's resilience in any period of fiscal retrenchment remained an open question.

The SingTel Innov8 contribution to Block 71 reflected a broader pattern in Singapore's innovation architecture: corporate-state hybrid investment vehicles. SingTel, majority-owned by Temasek, channelled its corporate VC activity through a programme that both served Singtel's strategic interests (access to startups building on telecoms infrastructure, potential acquisition targets) and served the state's interest in bootstrapping an ecosystem. Singtel Innov8 was launched in 2010 with a US$250 million fund; by 2025 it had backed more than 110 portfolio companies including SentinelOne, OpenTable and Arista, and total managed capital subsequently doubled to US$500 million with the addition of a US$250 million AI Growth Fund. The model was not without tension — portfolio companies sometimes found that corporate strategic objectives conflicted with the pure-financial-return logic that independent VCs applied — but as an ecosystem-seeding mechanism it served its purpose.

6. The Startup SG Programme — Founder, Tech, Equity, Talent

The Startup SG programme, launched formally in 2017 and operationalised through the newly formed Enterprise Singapore in 2018, was the government's most significant rationalisation of startup support since the T21 era. It brought under a single brand four distinct instruments that had previously operated through separate agencies with separate application processes, evaluation criteria, and reporting requirements.

Startup SG Founder provided a co-funding grant of up to S$50,000 to first-time entrepreneurs who had secured a Startup SG Accredited Mentor Partner (AMP) and demonstrated a viable business concept. The matching ratio operated at 3:1 (government to founder capital injection) at the programme's launch in March 2017; with effect from 1 April 2024, Enterprise Singapore revised the scheme to a 1:1 matching ratio over a 12-month period, with grant amounts of S$20,000–S$50,000, and expanded eligibility to allow first-time entrepreneurs to partner with non-first-time co-founders. The programme required mentorship as a condition of funding — an important design choice that aligned mentor incentives (accountability for their endorsed startups) with founder needs (access to experienced guidance). The programme's explicit focus on first-time entrepreneurs addressed a documented gap: Singapore's most successful founders (Sea's Forrest Li, Grab's Anthony Tan) had been exposed to foreign ecosystems before building their companies, but the programme aimed to create a comparable developmental pathway for founders who had not had international exposure.

Startup SG Tech provided non-dilutive grants for deep-tech startups commercialising intellectual property from Singapore's research institutions — Proof-of-Concept (POC) grants up to S$250,000 and Proof-of-Value (POV) grants up to S$500,000 per project at the scheme's launch; in 2025 these caps were raised by 60% to up to S$400,000 (POC) and S$800,000 (POV). The programme was designed to bridge the "valley of death" between research outputs and commercial products — the funding gap at which government research grants end and commercial VC begins. A*STAR spinouts, NUS technology licensing office ventures, and NTU Research commercialisation projects were primary beneficiaries. Deep-tech areas prioritised included advanced manufacturing, medical technology, clean technology, and AI — reflecting the government's sectoral economic priorities.

Startup SG Equity was the most capital-intensive pillar: the government co-invested alongside qualifying third-party investors in early-stage Singapore-registered technology companies through two mechanisms — the SEEDS Capital programme (under which co-investment is capped at S$2 million per general-tech startup and S$12 million per deep-tech startup as of 2025, applying 7:3 government-to-private ratios at the first institutional round before stepping down) and the Business Angel Scheme restructuring. The programme's logic was straightforward: by committing public co-investment alongside private lead investors, the government both reduced investors' individual risk and signalled government confidence in the quality of deal flow. The resulting portfolio became a de facto benchmark of Singapore startup quality. By 2026, nearly S$3 billion (combining more than S$2.5 billion in private-sector capital with public co-investment) had been invested in over 330 startups under the scheme, and Budget 2026 announced a further S$1 billion injection of government capital into Startup SG Equity.

Startup SG Talent addressed the supply-side constraint that Singapore's small domestic talent pool imposed on its startup ecosystem. The EntrePass scheme — a work visa specifically for foreign entrepreneurs seeking to incorporate and build businesses in Singapore — was streamlined and marketed under the Startup SG Talent umbrella. The programme also included the Tech@SG initiative (later absorbed into the Tech.Pass framework) which provided an employment-pass track for high-calibre technology professionals joining Singapore-based startups or establishing new ventures. By 2024, EntrePass and Tech.Pass holders contributed disproportionately to Singapore's AI and deep-tech startup cohorts .

The Startup SG programme architecture had notable strengths. Its branding provided a coherent public identity for what had been an opaque array of government schemes. Its AMPs created a distributed mentorship and quality-control layer that reduced Enterprise Singapore's direct assessment burden. Its four-pillar structure matched instruments to the specific market failures at each stage of startup development.

Its weaknesses were equally notable. The programme was optimised for compliance — for demonstrable grant utilisation, milestone achievement, and reporting — rather than for commercial success. Founders frequently described the administrative overhead of Singapore government grants as disproportionate to the quantum of funding. The AMP incentive structure created risks of capture, with some AMPs prioritising the facilitation of grant applications over genuine mentorship. And the programme's risk tolerance was lower than what a commercially oriented VC would apply: the preference for technically credible, low-risk propositions over transformative but uncertain bets meant that the most ambitious startups often found private capital more useful than government grants.

7. The Sovereign-Venture Architecture — Vertex, Heliconia, EDBI, Pavilion Capital

Singapore's sovereign-venture architecture — the network of state-linked investment vehicles operating in the startup and venture ecosystem — has no precise parallel in any other startup ecosystem. It is not the Israeli Yozma programme (a time-limited government-matching scheme), not the Chinese state VC model (which operates at far greater scale with less governance discipline), and not the Nordic public-VC model (which is smaller and more focused on domestic market development). It is a distinctively Singaporean instrument: patient, well-capitalised, governance-conscious, and calibrated to serve both commercial and strategic-economic objectives simultaneously.

Vertex Ventures, part of Temasek-owned Vertex Holdings, managed five successive Southeast Asia and India-focused venture funds — its Fund V closed in September 2023 at US$541 million, above its US$450 million target. Vertex Holdings as a group manages over US$6.8 billion across its global platform of partner funds in China, Southeast Asia and India, the US, Israel and Japan. Vertex's portfolio included Grab (early-stage investor), PropertyGuru, Shopee-era investments, and a growing cohort of AI and enterprise software companies. Vertex's competitive advantage relative to independent VCs was its Temasek parentage: access to Temasek's global network, the implicit endorsement of Singapore Inc., and the capacity to follow-on invest across multiple fund cycles without the capital-recycling pressure that finite commercial funds impose. Vertex became the most active state-linked VC in the region and, by reputation, one of the most disciplined — maintaining a return orientation rather than treating portfolio companies as policy instruments.

Heliconia Capital Management, established by Temasek in 2010, focused on growth-stage Singapore-based companies (not exclusively technology startups) with the mandate of helping Singapore SMEs scale. Its portfolio included companies in healthcare, food and beverage, and manufacturing, alongside technology-oriented businesses. Heliconia's role in the startup ecosystem was less prominent than Vertex's — its focus was on established SMEs rather than early-stage ventures — but its patient capital provided a bridge for Singapore-headquartered companies at the awkward growth stage between startup grant support and institutional PE or public markets.

EDBI, the corporate investment arm of the Economic Development Board, made strategic investments in technology companies on behalf of the Singapore government's industrial development objectives. Unlike Temasek's commercial mandate, EDBI's investments were explicitly strategic: the primary criterion was whether an investee company's technology, talent, or market position could strengthen Singapore's economic competitiveness or attract follow-on activity to Singapore. EDBI's portfolio by 2024 included AI infrastructure companies, semiconductor-adjacent ventures, and international deep-tech firms with Singapore R&D operations . The distinction between EDBI's strategic mandate and Vertex's commercial mandate was not always clean in practice, but the design intent differed.

Pavilion Capital, established in 2012 as a Temasek subsidiary focused on private equity in Southeast Asia, occupied the growth-stage and buyout end of the capital spectrum rather than venture. Its relevance to the startup ecosystem was as an exit vehicle: Pavilion provided a potential acquirer for Singapore-headquartered technology companies that had achieved scale but were not IPO-ready or that sought strategic buyers in preference to public markets. This exit-option contribution to the ecosystem is less visible than the investment function but equally important: founders and early investors can only recycle capital (and confidence) if viable exit paths exist.

Together, these sovereign-venture vehicles constituted what might be called a "patient capital stack" for Singapore's startup ecosystem. They provided capital at stages where commercial VCs were absent (EDBI at strategic technology stages, Heliconia at SME growth stages) and at scales that individual fund managers could not match (Vertex across five fund cycles). The architecture's critics argued that this patient capital crowded out the development of a genuine independent VC industry, kept alive companies that should have failed, and created a dependency culture in which Singapore founders oriented their businesses toward government-linked exit paths rather than genuine commercial competition. The government's response — that market failures in a small economy with limited private capital justify structural intervention — is the standard developmental-state justification, and the evidence for both sides remains contested.

8. The Singapore Unicorns — Grab, Sea, Carro, Carousell, Razer, NinjaVan, ShopBack

Singapore's eight unicorns (companies that achieved privately assessed valuations of US$1 billion or above) produced between 2014 and 2022 were diverse in sector, founder background, and capital structure. What they shared was Singapore as a headquarters jurisdiction — a regulatory anchor, a capital-raising node, and a talent hub — even as their revenue bases were geographically distributed across Southeast Asia.

Grab Holdings is the flagship of Singapore's unicorn generation. Founded in Malaysia in 2012 as MyTeksi by Anthony Tan and Tan Hooi Ling, Grab relocated its headquarters to Singapore in 2013. Its evolution from ride-hailing to food delivery, financial services (GrabPay, GrabFinance), and superapp architecture mirrored the ambitions of regional internet platforms globally. Grab received major SoftBank investment in its 2017 Series G; in early 2018 a US$2.5 billion round was jointly led by SoftBank and Didi at a US$6 billion pre-round valuation; in June 2018 Toyota led a US$1 billion round at a >US$10 billion valuation; and in March 2019 SoftBank Vision Fund injected US$1.46 billion, taking Grab's Series H total beyond US$4.5 billion and the company's valuation to about US$14 billion. Cumulative pre-IPO funding ran into the multiple billions of dollars across more than two dozen rounds. Its 2 December 2021 NASDAQ listing via SPAC merger with Altimeter Growth Corp valued Grab at approximately US$40 billion at deal close — at the time the largest SPAC merger on record and the biggest US debut by a Southeast Asian company. Shares fell more than 20% on their first trading day, and the share price declined further through 2022–2023 as the broader tech sell-off and ongoing losses prompted investor reassessment. For full-year 2024, Grab reported revenue of US$2.80 billion (up 19% year-on-year), narrowed its segment adjusted EBITDA loss to US$105 million (a 38% improvement), and announced its inaugural US$500 million share repurchase programme; Grab reached its first profitable quarter in early 2026.

Sea Limited (formerly Garena) is Singapore's most successful technology company by peak market capitalisation. Founded by Forrest Li in 2009 as an online gaming platform, Sea's tripartite structure — Garena (gaming), Shopee (e-commerce), and SeaMoney (digital financial services) — made it the defining Southeast Asian internet company of the 2010s. Sea's NYSE IPO on 20 October 2017 raised approximately US$884 million by selling 59 million American Depositary Shares at US$15 per ADS (priced above the indicative US$12–14 range, with no subsequent stock split). The share price climbed steeply through 2020 (up 385% on the year) and peaked at US$366.99 on 19 October 2021, taking Sea's market capitalisation close to US$200 billion and briefly making it the most valuable Southeast Asian company. The subsequent correction — as Garena gaming revenues contracted post-COVID and Shopee withdrew from multiple international markets — reduced Sea's market cap by more than 80% to around US$30 billion or lower through late 2022 and early 2023. By 2025, Sea had stabilised with narrower focus on its core Southeast Asian markets and a profitable Shopee and SeaMoney combination.

Razer Inc., founded by Min-Liang Tan and Robert Krakoff in 2005, was primarily a gaming peripherals hardware company that became a unicorn and listed on the Hong Kong Stock Exchange in 2017. While Razer had US origins and a global customer base, it maintained Singapore as its headquarters and was a prominent fixture of Singapore's technology identity narrative. Razer's inclusion in Singapore's unicorn roster is sometimes contested — its business was not built through Singapore's startup ecosystem in the same way as Grab or Carousell — but its HQ registration and Tan's Singapore citizen identity make it a standard inclusion.

Carousell was the most symbolically important of Singapore's unicorns because it was the most directly produced by Singapore's domestic ecosystem. Founded in 2012 by three NUS students (Marcus Tan, Lucas Ngoo, and Quek Siu Rong) who met through the NOC programme, Carousell grew from a mobile secondhand marketplace into a classifieds platform operating across Southeast Asia, Greater China, and South Asia. Its November 2019 merger with Telenor's 701Search produced a regional classifieds leader across eight markets in Southeast Asia, Hong Kong and Taiwan, lifting Carousell's valuation to over US$850 million; a subsequent US$80 million Naver-led round took valuation past US$900 million; and in September 2021 a US$100 million round led by South Korean PE firm STIC Investments valued Carousell at US$1.1 billion, formally crossing the unicorn threshold. Carousell's trajectory — NUS students, NOC exposure, Block 71 incubation, regional scaling, and eventual international partnership — was the template that Singapore's startup support infrastructure was designed to replicate.

Carro Automotive, founded by Aaron Tan in 2015, was Singapore's automotive marketplace and fintech unicorn, operating across Singapore, Indonesia, Thailand, Malaysia, and Japan. Carro's June 2021 Series C round (US$360 million led by SoftBank Vision Fund 2, with participation from MSIG and Indonesian-based EV Growth, at a valuation of just above US$1 billion) confirmed unicorn status. The company's AI-driven pricing and condition-assessment technology represented a deepening of the automotive e-commerce model. In April 2024 Carro was reported to be raising a US$100 million pre-IPO round at a >US$1.5 billion valuation, and by August 2025 was reported to be targeting a US listing at >US$3 billion — potentially the largest Southeast Asian US listing since 2017.

NinjaVan, founded by Lai Chang Wen in 2014, was Southeast Asia's leading tech-enabled logistics company, providing last-mile delivery for e-commerce across six countries. Its US$578 million Series E round (2021, co-led by GeoPost/DPDgroup) confirmed unicorn status. NinjaVan's growth tracked directly with the Shopee and Lazada e-commerce expansion, illustrating how Singapore's unicorn ecosystem was partly a function of regional platform growth rather than purely domestic innovation.

ShopBack, founded by Henry Chan and Joel Leong in 2014, was Singapore's cashback and rewards platform. It expanded across nine Southeast Asian markets, Australia, and South Korea. ShopBack's June 2022 Series F round of approximately US$80 million was reported at a pre-money valuation of around US$475 million — well short of the US$1 billion unicorn threshold that some earlier trade press had projected for the company. By 2024 ShopBack had pivoted toward AI-powered commerce recommendations, reflecting the AI wave's impact on existing platform companies. (ShopBack therefore sits in the "approached unicorn status" category rather than the confirmed-unicorn cohort.)

PatSnap (IP intelligence software) reached unicorn status with a US$300 million Series E led by SoftBank Vision Fund 2 and Tencent in March 2021. Advance Intelligence Group (parent of the Atome BNPL platform and ADVANCE.AI enterprise solutions) reached a >US$2 billion valuation with a US$400 million-plus Series D led by SoftBank Vision Fund 2 and Warburg Pincus in September 2021. FinAccel/Kredivo Holdings (BNPL and digital lending) was valued at approximately US$1.66 billion in October 2022 and closed a US$270 million Series D round led by Mizuho Bank in March 2023, confirming unicorn status.

The structural pattern across Singapore's unicorn cohort reveals an ecosystem that excelled at producing regional platform businesses with Singapore as headquarters and Southeast Asia as market. The notable absence — with partial exceptions in deep-tech — was globally competitive enterprise software, semiconductor design, or AI foundation model companies. This gap reflected the ecosystems' orientation toward consumer markets, Singapore's small domestic testing ground, and the longer gestation period of deep-tech ventures relative to the timeline of Block 71's founding.

9. The 2022–2024 Crypto Crisis — Three Arrows, FTX, Terra-Luna Fallout in Singapore

The 2022–2024 crypto crisis was the most significant test of Singapore's financial regulatory credibility since the 2002 corporate governance failures at Singapore-listed companies. Three distinct but interconnected implosions — Three Arrows Capital, Terra-Luna, and FTX Singapore — all had Singapore connections that the government could not easily explain away.

Three Arrows Capital (3AC), founded by Su Zhu and Kyle Davies, was headquartered in Singapore and incorporated in the British Virgin Islands. Singapore subsidiary Three Arrows Capital Pte. Ltd. was registered with MAS as a Registered Fund Management Company (RFMC) — a regime that capped AUM at S$250 million and limited the firm to no more than 30 qualified investors. At its peak (early 2022) 3AC reportedly managed approximately US$10 billion across leveraged cryptocurrency positions, well above any RFMC threshold. When the cryptocurrency market collapsed in May–June 2022, 3AC's leveraged positions — including a major exposure to the Luna/UST ecosystem — were wiped out. On 27 June 2022 the BVI court ordered 3AC into liquidation; the July 2022 bankruptcy filing listed approximately US$3.5 billion in creditor claims from 27 creditors, with Genesis Global Trading the largest at over US$2.3 billion. On 30 June 2022 MAS reprimanded Three Arrows Capital Pte. Ltd. for providing false information to MAS and exceeding the S$250 million AUM threshold (between July–September 2020 and November 2020–August 2021). In September 2023, MAS additionally issued nine-year Prohibition Orders against Zhu Su and Kyle Livingston Davies, prohibiting them from performing any regulated activity in Singapore.

Terra-Luna (the algorithmic stablecoin ecosystem) was co-founded by Do Kwon (Kwon Do-hyeong), a South Korean entrepreneur who had been running Terraform Labs from Singapore. The Terra-Luna collapse in May 2022 destroyed an estimated US$40–60 billion in aggregate market value globally — roughly US$40 billion in LUNA and US$18 billion in TerraUSD (UST) — making it one of the largest single financial implosions in cryptocurrency history. Do Kwon fled Singapore before international authorities could serve process; he was arrested at Podgorica airport in Montenegro in March 2023 carrying a falsified Costa Rican passport. Montenegro ruled in August 2024 to extradite him to South Korea (over the United States), but he was ultimately transferred to US custody and in December 2025 was sentenced to 15 years in US federal prison for what the sentencing court called "fraud on epic, generational scale".

FTX's Singapore-registered subsidiary Quoine Pte Ltd (which operated the Liquid exchange) was at the time of the November 2022 collapse exempt from holding a Payment Services Act licence while its DPT-service licence application was under MAS review — Quoine had not been granted, and was not at the point of collapse operating under, a Major Payment Institution licence. FTX.com itself was never licensed nor exempted in Singapore. When FTX's global operations collapsed in November 2022 following revelations of the commingling of customer and trading funds, MAS clarified publicly that FTX.com did not operate in Singapore but acknowledged that Singapore users had nonetheless been able to onboard directly to the offshore platform. Singaporean retail investors who had deposited assets with the various FTX entities faced uncertainty about asset recovery; recovery proceeded through US Chapter 11 bankruptcy and parallel Singapore winding-up actions for the local entities. MAS subsequently issued statements emphasising that being on the MAS watchlist or applying for licensing should not be read as endorsement of any digital-payment-token service provider's commercial operations.

The aggregate impact on Singapore's reputation was significant. Singapore had positioned itself as the preferred jurisdiction for regulated crypto activity in Asia, offering legal certainty (Payment Services Act), regulatory sandbox access, and an English-law contract environment. The concentration of major crypto failures among Singapore-connected entities undermined this positioning. International media coverage framed Singapore — sometimes unfairly — as a jurisdiction where regulatory arbitrage had been possible and oversight insufficient.

The government's response was calibrated and multi-pronged. MAS tightened the DPT service-provider licensing criteria substantially — raising capital requirements, tightening fit-and-proper tests for directors and controllers, and mandating customer-asset segregation requirements. MAS also increased its examination frequency for DPT licensees. The number of entities holding or applying for DPT-service licences contracted sharply from the 2021 application peak to a much smaller cohort of active licensees by 2024 . MAS simultaneously signalled continued openness to "responsible digital asset innovation" — particularly in tokenisation of real-world assets and wholesale digital finance — while explicitly distancing this from speculative retail cryptocurrency.

The crypto crisis had an indirect but important effect on the broader startup ecosystem. It accelerated the pivot of government policy attention and ecosystem funding toward AI and deep-tech. Enterprise Singapore and EDB messaging in 2023–2024 increasingly foregrounded AI, precision medicine, and advanced manufacturing rather than fintech or digital payments. Several fintech-era accelerators and corporate innovation labs that had focused heavily on crypto-adjacent businesses retooled or closed. The surviving fintech ecosystem was smaller, more focused on enterprise payments and embedded finance, and more oriented toward government procurement.

10. The 2024–2026 AI Startup Wave

The AI startup wave that emerged in Singapore between 2024 and 2026 was qualitatively distinct from previous waves in the ecosystem's history. Earlier waves — consumer internet (2011–2016), fintech (2016–2020), deep-tech and biotech (2018–2022) — had all been characterised by adaptation and implementation: Singapore startups applying globally generated technology frameworks (mobile internet, blockchain, CRISPR) to Southeast Asian market opportunities. The AI wave, by contrast, saw Singapore startups attempting to participate at the frontier of technology development, not only in application.

This shift was enabled by three infrastructure investments that matured simultaneously around 2023–2024. First, the National AI Strategy 2.0 — launched on 4 December 2023 — was backed in Budget 2024 by an investment of more than S$1 billion over five years to support its implementation, including up to S$500 million dedicated to securing high-performance compute resources for AI innovation, dramatically reducing the capital barrier for Singapore-based AI researchers and founders. Complementary tranches followed, including a S$150 million Enterprise Compute Initiative and a further S$120 million for "AI for Science" led by the National Research Foundation. Second, AI Singapore (AISG), established in 2017 and expanded in 2020–2022, had by the mid-2020s engaged more than 700 companies and 15,000 professionals across its outreach and industry programmes, trained roughly 280–400 Singaporean AI engineers via its AI Apprenticeship Programme (AIAP) with a near-100% placement rate, and supported a growing pipeline of 100 Experiments (100E) projects. Third, the concentration of hyperscaler data centres in Singapore — AWS, Google Cloud, Microsoft Azure, and Oracle all operated Singapore-region infrastructure — provided the commercial cloud infrastructure on which AI workloads could be deployed.

The AI startups emerging in this wave fell into several categories. Foundation-model fine-tuning ventures — companies building Southeast Asia-specific or domain-specific language models on top of open-source foundations (Llama, Mistral, Falcon) — proliferated as the barrier to training domain-adapted models fell. Several ASTAR Institute for Infocomm Research (I²R) spinouts and NUS School of Computing affiliated ventures fell into this category [TBD-VERIFY: a named, public list of ASTAR I²R / NUS SoC AI spinouts incorporated 2024–2026 is not consolidated in public sources; requires A*STAR Innovation/NUS ILO commercialisation registers]. Enterprise AI deployment companies — systems integrators and SaaS providers wrapping AI capabilities for specific verticals (legal, healthcare, logistics, financial services) — represented the largest segment by headcount and revenue. AI-native vertical SaaS companies — building products in which AI was the core value proposition rather than a feature — were the fastest-growing segment by deal count. And a small but growing cohort of AI infrastructure and tooling companies (vector databases, AI observability, model evaluation platforms) served the developer and enterprise market globally from Singapore bases.

The government's role in the AI wave was again structurally enabling rather than passive. Enterprise Singapore's Startup SG programme added AI-specific pathways. EDB's investment promotion work targeted AI-adjacent MNC R&D centres, which in turn seeded corporate venture activity. The NUS-Industry Collaboration Programme and A*STAR's commercialisation office expanded their deep-tech translation resources with AI focus. And the establishment of Singapore's AI governance infrastructure — the Model AI Governance Framework (2nd edition, 2020), the AI Verify testing toolkit (2022), and the forthcoming AI Safety and Governance Centre — provided the regulatory clarity that enterprise AI buyers required before deploying AI-powered products from Singapore-based vendors.

By mid-2026, the AI startup wave had not yet produced a new unicorn, but the pipeline was visible: several well-funded AI-native companies had achieved Series A or B valuations in the S$200–500 million range, and at least two were reported to be in pre-IPO discussions . The wave also had a pronounced talent dimension: Singapore's universities were producing AI-trained graduates at higher rates than any previous technology wave, and the EntrePass and Tech.Pass frameworks were attracting AI-specialist founders from India, China, and the United States who found Singapore's regulatory environment and capital access attractive.

11. Comparative Lens — Singapore vs Tel Aviv, Bangalore, Shanghai, Stockholm

Comparing Singapore's startup ecosystem against its most cited peers illuminates both its distinctive strengths and its structural constraints. The four most instructive comparators are Tel Aviv (innovation intensity per capita), Bangalore (talent volume and services depth), Shanghai (state-capital integration), and Stockholm (per-capita unicorn production from a small domestic market).

Singapore and Tel Aviv are the most frequently paired comparators, and the pairing is instructive precisely because the surface similarities — small domestic market, export-dependent economy, heavy state investment in R&D — obscure fundamental structural differences. Tel Aviv's deep-tech intensity is rooted in military technology transfer: the IDF's elite technology units (Unit 8200, Unit 81) produce cybersecurity, signals intelligence, and systems engineering alumni at a rate that no civilian programme can replicate. The cultural dimension is equally important — Israeli entrepreneurial culture, shaped by national service, existential threat, and a flat hierarchy that normalises disputing authority, produces a risk appetite and a tolerance for failure that Singapore's meritocracy-oriented culture does not naturally generate. Israel's venture capital industry, seeded by the Yozma programme (1993–2000) and sustained by diaspora networks in Silicon Valley and New York, operates with a commercial independence that Singapore's government-co-investment model moderates. Singapore scores better than Israel on regulatory quality, English-language accessibility, and Southeast Asian market connectivity, but Israel's per-capita output of cybersecurity, AI research, and semiconductor design startups remains higher.

Singapore and Bangalore represent fundamentally different startup ecosystem models. Bangalore's scale — approximately 12,000 active startups, a talent pool of over 1.5 million engineers, and domestic market access to 1.4 billion Indian consumers — is categorically different from Singapore's. Bangalore's AI and SaaS ecosystem, fed by IIT and IISc graduate pipelines, produces a volume of enterprise software companies that Singapore cannot match. Singapore's advantages relative to Bangalore are governance quality (contract enforcement, IP protection, regulatory predictability), capital market access (SGX, international VC concentration), and ASEAN market gateway positioning. The two ecosystems are, in practice, increasingly complementary rather than competing: Singapore-headquartered companies (including several of the AI-wave startups) are building technology with Bangalore engineering teams and distributing products through Singapore's regulatory infrastructure.

Singapore and Shanghai share a developmental-state orientation but operate under fundamentally different political economies. Shanghai's Zhangjiang Hi-Tech Park and the Yangtze River Delta ecosystem benefit from China's domestic market of 1.4 billion consumers, CCP-directed industrial policy with far greater resource deployment than Singapore can marshal, and a VC ecosystem (Sequoia China, Hillhouse, GGV) operating at scale orders of magnitude larger than Singapore's. The 2021–2022 Chinese regulatory crackdown on technology platforms — which destroyed significant value in Alibaba, Didi, Tencent, and the broader tech ecosystem — created a temporary window in which Singapore positioned itself as a more stable alternative for Chinese tech talent and capital. This window drove a significant influx of Chinese tech founders, engineers, and VC fund managers to Singapore in 2022–2023, materially enriching the ecosystem's talent density. Indicators included ByteDance founder Zhang Yiming using Singapore as a primary base from 2022; ByteDance transferring roughly 1,000 chip engineers to a Singapore-incorporated subsidiary (Picoheart) from late 2023; and the relocation of multiple Chinese-origin crypto and AI entrepreneurs to Singapore in this period .

Singapore and Stockholm are the most precisely comparable in ecosystem structure. Sweden has a population of approximately 10 million (versus Singapore's 5.9 million), a similar openness to trade and foreign capital, a strong engineering education base, and a cultural premium on innovation. Stockholm's unicorn production — Spotify, Klarna, King, Mojang, iZettle, Truecaller, Northvolt, and others — is extraordinary on a per-capita basis, arguably the highest of any city outside the United States. The structural reasons are instructive: Sweden's flat organisational hierarchy reduces the career risk of leaving a large employer to found a startup; the Swedish social safety net reduces the personal cost of failure; early digitisation of music and gaming industries created a generation of software entrepreneurs; and a strong culture of design has driven product quality beyond what pure engineering prowess produces. Singapore has imported elements of the Stockholm model (the flat aesthetic of Block 71, the emphasis on product design at NUS, government safety nets via ComCare for failed entrepreneurs) but has not replicated the underlying cultural conditions that generate Stockholm's output.

12. Conclusion

Singapore's startup ecosystem, assessed in 2026, is both an achievement and a work in progress. The achievement is substantial: a fifteen-year construction, from nothing more than repurposed JTC industrial buildings and a handful of government programmes, has produced eight unicorns, hundreds of millions of dollars in venture capital deployed, thousands of technology companies, and a pipeline of AI-native founders that did not exist a decade ago. By the metrics that matter to economic policymakers — employment created, revenue generated, investment attracted, and technologies commercialised — the investment in startup ecosystem infrastructure has delivered positive returns.

The work in progress is equally substantial. The structural constraints that Singapore's planners identified in the 2010 ESC report — risk aversion, insufficient tolerance for failure, thin private VC industry, small domestic market — have not been eliminated by fifteen years of ecosystem investment. They have been mitigated at the margin: Block 71 created a physical community that normalised entrepreneurship; Startup SG grants reduced the personal financial cost of founding; sovereign-venture vehicles provided capital at stages where private capital was absent; and the NUS Overseas Colleges and equivalent programmes produced a cohort of founders with international exposure. But the underlying conditions — a labour market that rewards stable MNC employment, a housing cost structure that makes financial failure catastrophic, and a meritocracy ideology in which failure carries reputational cost — remain.

The 2022–2024 crypto crisis administered a corrective that, in retrospect, may prove to have been productive. By forcing a reckoning with the limits of light-touch regulatory tolerance for speculative digital-asset activity, the crisis accelerated Singapore's reorientation toward AI and deep-tech — sectors in which the developmental state's comparative advantage (patient capital, public R&D infrastructure, talent pipeline investment) is more clearly deployable than in consumer-internet or speculative finance. The AI wave of 2024–2026, still early in its trajectory, represents a more defensible long-term economic position than the fintech maximalism of 2018–2022.

Singapore's startup ecosystem in 2026 remains a managed ecosystem — managed by design, by resource allocation, and by the deliberate choices of state agencies that shape what gets built, where, and by whom. The question is not whether management is necessary (for a city-state of 5.9 million with no natural technology hinterland, it is) but whether the management is calibrated correctly for the next phase of the technology economy. The AI era requires different infrastructure, different skills, and different risk tolerances than the consumer-internet era that Block 71 was designed to seed. Whether Singapore's ecosystem can adapt — whether it can produce companies that lead at the technology frontier rather than adapt frontier technologies for regional markets — is the central question of the next decade.

13. Spiral Index

This document connects to the following thematic axes within the corpus:

Developmental state and directed economic transformation: See SG-M-09 (Developmental State — Singapore's Variant), SG-E-46 (Industrial Strategy), SG-E-27 (Committee on the Future Economy), SG-E-01 (Economic Development Board)
Technology and digital economy infrastructure: See SG-E-25 (Digital Economy), SG-D-17 (Technology and Smart Nation), SG-O-01 (AI Mega Trend), SG-O-12 (AI Governance Deep-Dive)
Research and innovation ecosystem: See SG-E-15 (Research, Innovation, and Enterprise), SG-E-16 (A*STAR)
Crypto, fintech, and financial regulation: See SG-E-36 (Crypto, Fintech, and Family Office Hub)
Labour market and AI displacement: See SG-O-14 (Jobs Versus AI in Singapore), SG-E-47 (Wage Models — IWI, PWM, and WCS)
Lawrence Wong era and Forward Singapore: See SG-B-09 (Lawrence Wong Transition)

Sources cited above. For primary-source verification of items marked [TBD-VERIFY], consult: NUS Enterprise annual impact reports; Enterprise Singapore Startup SG programme statistics (published annually); Grab Holdings and Sea Limited SEC filings; MAS enforcement register; DealStreetAsia and e27 ecosystem databases; Startup Genome Global Startup Ecosystem Report (annual).

SG-E-49: The Singapore Startup Ecosystem — From Block 71 to the AI Era (2010–2026)