Published 2026-08-21 | Version v1.0
Working PaperOpenPublished

Toward Measuring AI Infrastructure Investment and Economic Resilience Across Ten Economies

Financing Architectures, Capital Formation, and Deployment Timing

Description

This working paper develops a measurement framework for comparing AI infrastructure investment and economic-resilience context across ten economies. It argues that AI infrastructure financing cannot be assessed through headline investment totals alone because records differ in realization status, AI specificity, accounting basis, capital-formation eligibility, national coverage, and deployment timing. The study compares financing architectures in the United States, United Kingdom, Canada, Japan, South Korea, Germany, France, Singapore, China, and India, while treating AI-IIS, PAIIS, and PrAIIS as bounded measurement concepts rather than presently comparable national statistics. It concludes that meaningful cross-country investment comparison and later resilience testing require aligned scope, period matching, denominator completeness, and operational-capacity timing.

Abstract

Countries are expanding AI infrastructure through markedly different combinations of public and private capital, yet cross-country assessment is complicated by inconsistent investment accounting, incomplete national coverage, and the lag between expenditure and usable capacity. This study compares financing architectures across ten economies: the United States, United Kingdom, Canada, Japan, South Korea, Germany, France, Singapore, China, and India. Public AI Infrastructure Investment Share (PAIIS), Private AI Infrastructure Investment Share (PrAIIS), and AI infrastructure investment intensity (AI-IIS) are treated as bounded measurement concepts rather than presently comparable national statistics. The analysis distinguishes source quality, realization status, AI specificity, accounting basis, capital-formation eligibility, and deployment timing. Qualifying AI-specific realized monetary records are available from five economies, but national coverage remains incomplete and several records combine capital and operating expenditure. France provides completed-project funding with unresolved payment timing; Germany and Singapore provide procurement evidence; China provides realized broader-compute CAPEX that is too broad for an AI-only numerator; and India reports mission-wide expenditure without an infrastructure-specific split. The study therefore compares financing architecture, monetary evidence, and deployment timing rather than constructing national investment rankings. Macroeconomic and labor-market indicators are reported only as descriptive context. The findings show that scope, accounting treatment, realization status, deployment timing, and coverage must be aligned before meaningful cross-country investment comparisons or subsequent resilience analysis can be undertaken.

Files

PDF preview
PDF preview

Mobile browsers may not display embedded PDF previews reliably. Open the PDF directly for the best reading experience.

Open PDF Download PDF
Files
NameType
Toward Measuring AI Infrastructure Investment and Economic Resilience Across Ten Economies Financing Architectures, Capital Formation, and Deployment Timing.pdf
Full-text PDF of the working paper		application/pdfDownload

Keywords

  • AI infrastructure investment
  • public-private financing
  • investment measurement
  • AI compute
  • capital formation
  • deployment lag
  • cross-country comparison
  • economic resilience
  • PAIIS
  • PrAIIS
  • AI-IIS
  • public capital
  • private investment
  • AI infrastructure
  • data centers
  • GPU programs
  • AI compute capacity
  • measurement framework
  • deployment timing
  • EPINOVA

Subjects

  • Artificial intelligence infrastructure
  • Investment measurement
  • Economic resilience
  • Public-private financing
  • Digital infrastructure
  • Capital formation
  • AI compute
  • Industrial policy
  • Comparative political economy
  • Public policy
  • Economics
  • Technology governance

Recommended citation

Wu, Shaoyuan. (2026). Toward Measuring AI Infrastructure Investment and Economic Resilience Across Ten Economies: Financing Architectures, Capital Formation, and Deployment Timing (EPINOVA Working Paper No. EPINOVA–WP–D–2026–03). Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.67037/epinova.wp.d.2026.003

APA citation

Wu, S. (2026). Toward measuring AI infrastructure investment and economic resilience across ten economies: Financing architectures, capital formation, and deployment timing. EPINOVA Working Paper Series, EPINOVA-WP-D-2026-003. Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.67037/epinova.wp.d.2026.003.

Alternate identifiers

SchemeIdentifierDescription
URLhttps://epinova.org/working-papersOfficial EPINOVA working papers page
EPINOVA working paper numberEPINOVA–WP–D–2026–03Working paper number printed in the PDF
File nameToward Measuring AI Infrastructure Investment and Economic Resilience Across Ten Economies Financing Architectures, Capital Formation, and Deployment Timing.pdfSource PDF file name
Measurement conceptAI-IISAI infrastructure investment intensity, valid only when numerator and GDP denominator periods and capital-formation scope are aligned
Measurement conceptPAIISPublic AI Infrastructure Investment Share for bounded financing pools with defensible total-investment denominators
Measurement conceptPrAIISPrivate AI Infrastructure Investment Share for bounded financing pools with defensible total-investment denominators

Related works

RelationIdentifierTypeDescription
IsPartOfhttps://epinova.org/working-papersPublication seriesEPINOVA Working Paper Series
IsSupplementedByhttps://github.com/EPINOVALLC/EPINOVA-ResearchRepositorySupplementary repository and structural archive
ReferencesBrynjolfsson, E., Rock, D., & Syverson, C. (2021). The productivity J-curveJournal articleReferenced for the deployment-lag and complementary-investment logic
ReferencesBresnahan, T. F., & Trajtenberg, M. (1995). General purpose technologiesJournal articleReferenced for AI as a general-purpose technology
ReferencesSimmie, J., & Martin, R. (2010). The economic resilience of regionsJournal articleReferenced for evolutionary economic-resilience framing
ReferencesSoftBank Corp. (2026). FY2025 full-year investor briefingCorporate reportReferenced for Japanese AI computing and AI data-center CAPEX evidence
ReferencesNational Assembly Budget Office. (2026). Government GPU purchase programGovernment accounts analysisReferenced for South Korea GPU-program execution and deployment timing

References

  1. Bom, P. R. D., & Ligthart, J. E. (2014). What have we learned from three decades of research on the productivity of public capital? Journal of Economic Surveys, 28(5), 889–916. https://doi.org/10.1111/joes.12037
  2. Bresnahan, T. F., & Trajtenberg, M. (1995). General purpose technologies ‘engines of growth’? Journal of Econometrics, 65(1), 83–108. https://doi.org/10.1016/0304-4076(94)01598-T
  3. Bristow, G., & Healy, A. (2018). Innovation and regional economic resilience: An exploratory analysis. The Annals of Regional Science, 60(2), 265–284. https://doi.org/10.1007/s00168-017-0841-6
  4. Brynjolfsson, E., Rock, D., & Syverson, C. (2021). The productivity J-curve: How intangibles complement general purpose technologies. American Economic Journal: Macroeconomics, 13(1), 333–372. https://doi.org/10.1257/mac.20180386
  5. China Mobile Limited. (2025, March 20). 2024 annual results presentation. https://www.chinamobileltd.com/en/ir/webcasts/pre250320.pdf
  6. China Telecom Corporation Limited. (2026, March 24). 2025 annual results. https://www.chinatelecom-h.com/en/ir/presentations/annpre260324.pdf
  7. CNRS. (2024, March 28). GENCI and CNRS choose Eviden to make the Jean Zay supercomputer one of the most powerful in France. https://www.cnrs.fr/en/press/genci-and-cnrs-choose-eviden-make-jean-zay-supercomputer-one-most-powerful-france
  8. de Vries, A. (2023). The growing energy footprint of artificial intelligence. Joule, 7(10), 2191–2194. https://doi.org/10.1016/j.joule.2023.09.004
  9. Engel, E., Fischer, R., & Galetovic, A. (2013). The basic public finance of public-private partnerships. Journal of the European Economic Association, 11(1), 83–111. https://doi.org/10.1111/j.1542-4774.2012.01105.x
  10. EuroHPC Joint Undertaking. (2023, October 3). Procurement contract for JUPITER, the first European exascale supercomputer, signed. https://www.eurohpc-ju.europa.eu/procurement-contract-jupiter-first-european-exascale-supercomputer-signed-2023-10-03_en
  11. GENCI. (2024, November 20). New milestone for French AI flagship Jean Zay: H100 extension. https://www.genci.fr/en/news/new-milestone-french-ai-flagship-jean-zay-drive-major-ai-breakthroughs-thanks-record-eviden
  12. Government of India, Ministry of Finance. (2026). Expenditure profile 2026–2027: Statement 6. https://www.indiabudget.gov.in/doc/eb/stat6.pdf
  13. Guo, P. (2025). Digital infrastructure and firm resilience: Evidence from China. Emerging Markets Finance and Trade, 61(5), 1342–1359. https://doi.org/10.1080/1540496X.2024.2416478
  14. International Monetary Fund. (2026, April 14). World Economic Outlook, April 2026: Statistical Appendix A tables. https://www.imf.org/-/media/files/publications/weo/2026/april/english/tablea.pdf
  15. Juhasz, R., Lane, N., & Rodrik, D. (2024). The new economics of industrial policy. Annual Review of Economics, 16, 213–242. https://doi.org/10.1146/annurev-economics-081023-024638
  16. Kydland, F. E., & Prescott, E. C. (1982). Time to build and aggregate fluctuations. Econometrica, 50(6), 1345–1370. https://doi.org/10.2307/1913386
  17. Lannelongue, L., Grealey, J., & Inouye, M. (2021). Green algorithms: Quantifying the carbon footprint of computation. Advanced Science, 8(12), 2100707. https://doi.org/10.1002/advs.202100707
  18. Martin, R. (2012). Regional economic resilience, hysteresis and recessionary shocks. Journal of Economic Geography, 12(1), 1–32. https://doi.org/10.1093/jeg/lbr019
  19. Martin, R., & Sunley, P. (2015). On the notion of regional economic resilience: Conceptualization and explanation. Journal of Economic Geography, 15(1), 1–42. https://doi.org/10.1093/jeg/lbu015
  20. Matvejevs, O., & Tkacevs, O. (2025). Invest one - get two extra: Public investment crowds in private investment. European Journal of Political Economy, 90, 102384. https://doi.org/10.1016/j.ejpoleco.2023.102384
  21. Meta. (2025). Hello, Beaver Dam. Meta Data Centers. https://datacenters.atmeta.com/2025/11/hello-beaver-dam/
  22. Ministry of Economy, Trade and Industry, Japan. (2024). Approval of cloud programs for AI computing resources. https://www.meti.go.jp/english/press/2024/0419_001.html
  23. Ministry of Industry and Information Technology of China. (2025, May 30). Computing power interconnection action plan. https://app.www.gov.cn/govdata/gov/202505/30/530146/articleNew.html
  24. National Assembly Budget Office. (2026). 2025 accounts analysis: Science and ICT Committee - government GPU purchase program. https://nabo.go.kr/board/file/down.do?fid=33319407
  25. Noy, S., & Zhang, W. (2023). Experimental evidence on the productivity effects of generative artificial intelligence. Science, 381(6654), 187–192. https://doi.org/10.1126/science.adh2586
  26. NSCC Singapore. (2026, June 8). NSCC Singapore launches next-generation supercomputer ASPIRE 2B. https://www.nscc.sg/portfolio/item/nscc-singapore-launches-next-generation-supercomputer-aspire-2b/
  27. OECD. (2026a, April 16). Labour market situation: Employment and labour force participation remained stable in most OECD countries in Q4 2025. https://www.oecd.org/en/data/insights/statistical-releases/2026/04/labour-market-situation-updated-april-2026.html
  28. OECD. (2026b). OECD compendium of productivity indicators 2026. OECD Publishing. https://doi.org/10.1787/734a5e68-en
  29. Parliament of India, Ministry of Electronics and Information Technology. (2026, March). IndiaAI Mission expenditure [Rajya Sabha parliamentary answer]. https://sansad.in/getFile/annex/270/AU3246_wa5YLu.pdf?source=pqars
  30. Public Services and Procurement Canada. (2025). Public Accounts of Canada 2025, Volume II: Innovation, Science and Industry transfer payments. https://www.tpsgc-pwgsc.gc.ca/recgen/cpc-pac/2025/vol2/isi/pt-tp-eng.html
  31. RecordOwl. (2026). Hewlett-Packard Singapore sales Pte. Ltd.: Government procurement record [GeBIZ mirror]. https://recordowl.com/company/hewlett-packard-singapore-sales-pte-ltd
  32. SAKURA internet Inc. (2026, April 27). Fiscal year ended March 2026 financial results. https://www.sakura.ad.jp/corporate/wp-content/uploads/2026/04/en-260427-ir_2.pdf
  33. Sensier, M., Bristow, G., & Healy, A. (2016). Measuring regional economic resilience across Europe: Operationalizing a complex concept. Spatial Economic Analysis, 11(2), 128–151. https://doi.org/10.1080/17421772.2016.1129435
  34. Sevilla, J., Heim, L., Ho, A., Besiroglu, T., Hobbhahn, M., & Villalobos, P. (2022). Compute trends across three eras of machine learning. 2022 International Joint Conference on Neural Networks (IJCNN), 1–8. https://doi.org/10.1109/IJCNN55064.2022.9891914
  35. Simmie, J., & Martin, R. (2010). The economic resilience of regions: Towards an evolutionary approach. Cambridge Journal of Regions, Economy and Society, 3(1), 27–43. https://doi.org/10.1093/cjres/rsp029
  36. SoftBank Corp. (2026, May 11). FY2025 full-year investor briefing: Status of investments in AI computing infrastructure and AI data centers. https://www.softbank.jp/en/corp/set/data/ir/documents/presentations/fy2025/investors/pdf/sbkk_investors_presentation_20260511_en.pdf
  37. UK Research and Innovation. (2025). Annual report and accounts 2024 to 2025. https://www.ukri.org/publications/annual-report-and-accounts-2024-to-2025/ukri-annual-report-and-accounts-2024-to-2025/
  38. Wisconsin Legislative Fiscal Bureau. (2026). Sales tax exemption for WEDC certified data centers [Memorandum]. https://www.wpr.org/wp-content/uploads/2026/04/18-Hesselbein-Habush-Sinykin-ST-JG-Adobe-cloud-storage.pdf
  39. World Bank. (2026a). Employment-to-population ratio, ages 15+, total (%) (modeled ILO estimate). Retrieved August 19, 2026, from https://data.worldbank.org/indicator/SL.EMP.TOTL.SP.ZS
  40. World Bank. (2026b). Unemployment, youth total (% of total labor force ages 15–24) (modeled ILO estimate). Retrieved August 19, 2026, from https://data.worldbank.org/indicator/SL.UEM.1524.ZS
  41. Zhang, J., Yang, Z., & He, B. (2023). Does digital infrastructure improve urban economic resilience? Evidence from the Yangtze River Economic Belt in China. Sustainability, 15(19), 14289. https://doi.org/10.3390/su151914289