Home / Repository / FR / FR_WLD_2026_706
fr

Reproducibility package for World Development Report 2026: The Promise of Artificial Intelligence

2026
Reference ID
FR_WLD_2026_706
DOI
https://doi.org/10.60572/6n9m-dk62
Author(s)
Gaurav Nayyar, Anukriti Siddharth, Sharmista Appaya, Elwyn Davies, Lelys Dinarte-Diaz, Sam Fraiberger, Yi Jie Gwee, Yan Liu, Jeremy Ng, Tiago C. Peixoto, Manuel Ramos-Maqueda, Anja Sautmann, Katherine Stapleton, Shu Yu
Collections
Flagships and Reports
Metadata
JSON
Created on
Aug 11, 2026
Last modified
Aug 14, 2026
Page views
153
Downloads
28
  • Overview
  • Reproducibility Package
  • Description
  • Scope and coverage
  • Disclaimer
  • Access and rights
  • Contacts
  • Information on metadata
  • Citation

    • Overview

    Abstract

    World Development Report 2026 will investigate the development implications of AI as a general-purpose technology (GPT) and assess what might be the best policy choices to leverage the benefits of AI while offsetting potential risks. In doing so, it will focus on the institutional and governance arrangements needed to ensure inclusive and responsible deployment of AI. The aim of the Report is to provide a developing-country perspective on a topic for which the academic and policy discussion has focused primarily on high-income countries.

    Reproducibility Package

    Scripts
    File name
    FR_WLD_2026_706
    Zip package
    FR_WLD_2026_706.zip
    Title
    Reproducibility package for World Development Report 2026: The Promise of Artificial Intelligence
    Date
    2026-08
    Dependencies
    R dependencies are listed in the file renv.lock. Stata dependencies are listed in the ado folder. Python dependencies are listed in requirements.txt.
    Instructions
    See README in reproducibility package.
    Notes
    Computational reproducibility verified by Development Impact (DECDI) Analytics team, World Bank.
    Source code repository
    Repository name URI
    Reproducible Research Repository (World Bank) https://reproducibility.worldbank.org
    Software
    Name
    R
    Version
    4.5.1
    Name
    Stata
    Version
    19.5 MP
    Name
    Excel

    Reproducibility

    Technology environment

    Report exhibits were reproduced on computers with the following specifications:

    1. Computer 1: Chapter 4

      • OS: Windows 11 Enterprise
      • Processor: Intel(R) Xeon(R) Platinum 8562Y+ 2.80 GHz (2 processors)
      • Memory available: 32.0 GB
      • Software versions: Stata 19.5 MP; R 4.6.1

    2. Computer 2: Chapter 1, Spotlight 7, WBES AI Replication Package

      • OS: Windows 11 Enterprise
      • Processor: Intel(R) Xeon(R) Platinum 8562Y+ 2.80 GHz (2 processors)
      • Memory available: 32.0 GB
      • Software versions: Stata 19.5 MP; R 4.5.3

    3. Computer 3: Chapter 3, Chapter 6, Chapter 9

      • OS: Windows 11 Enterprise 25H2
      • Processor: Intel(R) Xeon(R) Gold 6226R CPU @ 2.90 GHz (2 processors)
      • Memory available: 16.0 GB
      • Software versions: R 4.6.1 (run in RStudio)

    4. Computer 4: Chapter 8

      • OS: Windows 11 Enterprise
      • Processor: Intel(R) Xeon(R) Platinum 8562Y+ 2.80 GHz (2 processors)
      • Memory available: 32.0 GB
      • Software versions: R 4.6.1; Stata 19.5 MP

    5. Computer 5: Overview, Chapter 2, Chapter 5, Chapter 7, Spotlight 3, Spotlight 5

      • OS: macOS Tahoe 26.6
      • Processor: Apple M4 Pro
      • Memory available: 24.0 GB
      • Software versions: Stata 19.5 MP; R 4.6.1
    Technology requirements

    Runtime: ~40 minutes
    Reproduction instructions

    The package is organized as one sub-package per chapter/spotlight, each with its own README describing data sources, code, and execution steps; a consolidated README at the root summarizes the full package. Most figures start from the raw or intermediary datasets included in each chapter's "1 - Data" folder; several figures are produced directly in Excel with no code involved.

    A new user will be able to reproduce the findings of the report by doing the following:

    Chapter 1 (runtime: 2 minutes)

    1. Opening the Rproj under the Chapter 1 folder.
    2. Updating the project globals and running the Stata fig 1.2 do file.

    Chapter 2 (runtime: 4 minutes)

    1. Opening the Rproj under the Chapter 2 folder and running the R script.
    2. Updating the project globals and running the Stata ffigures_2_2_2_5 do file.

    Chapter 3 (runtime: 2 minutes)

    1. Opening the Rproj under the Chapter 3 folder and running the R scripts (Figure_3.1_a, Figure_3.1_b).

    Chapter 4

    1. Figures prepared in Excel (no code involved): Figures 4.3a, 4.6, and 4.9.
    2. Figure 4.3b: Updating the working directory in the do-file Chapter 4/2 - Code/Figure 4.3 panel b.
    3. Figure B4.2.1: Updating the working directory in the do-file Figure B4.2.1 and the R script Figure B4.2.1.Rmd.
    4. Figure 4.7: Updating the working directory of the R script Chapter 4/2 - Code/Figure_4_7/03_make_figure_4_7.
    5. Figure B4.3.1: Updating the working directory in the do-file Chapter 4/2 - Code/Code_B4.3.1/0master_do_file.
    6. Figure 4.8: Updating the directory in the code files in Chapter 4/2 - Code/Figure 4.8.
    7. Figure 4.10a: Updating the working directory in the R script Figure_4_10_panel_a/prepare_data.
    8. Figure 4.12a: Updating the working directory in the do-file Figure_4_12_panel_a.

    Chapter 5 (runtime: 5 minutes)

    1. Adjusting the file paths in Chapter 5 Full R Reproducibility.Rmd and running the script.
    2. Adjusting the file paths in the Radiologists Replication do file to point to the local copy of the package, and running the do file.

    Chapter 6 (runtime: 2 minutes)

    1. Opening the .Rproj, opening main.R, and running the code.

    Chapter 7 (runtime: 2 minutes)

    1. Opening the do-file Table_7_1, changing the path, and running it.
    2. Reviewing the Excel-based figures.

    Chapter 8 (runtime: 3 minutes)

    1. Opening the R project Chapter 8.
    2. Restoring the environment by running renv::restore() and following the prompts.
    3. Opening Chapter 8 Full R Reproducibility.Rmd and running the code.

    Chapter 9 (runtime: 2 minutes)

    1. Opening the Chapter 9.Rproj and running the main R Script.

    Spotlight 3 (runtime: 1 minute)

    1. Updating the working directory on line 1 of the TALIS replication do file and running it.

    Spotlight 4

    1. Figures prepared in Excel (no code involved).

    Spotlight 5 (runtime: 1 minute)

    1. Updating the working directory in line 20 of the do-file Figure S_5_1 replication and running it.

    Spotlight 7 (runtime: 2 minutes)

    1. Opening the R project Spotlight 7.Rproj.
    2. Restoring the environment by running renv::restore(), and following the prompts.
    3. Opening AI_Agency_Income_Analysis.Rmd and running the code.

    WBES AI Replication Package (runtime: 5 minutes)

    1. Updating the working directory at the beginning of each do-file and running aies_replication_run do.

    Since not all the data is included, the package includes the results produced by replicators. These files can be used to review the results presented in the report.

    Data

    Datasets
    Name
    Generative AI Adoption Data
    Note
    Source: Semrush web traffic data on generative AI (proprietary, subscription-based), extracted via Semrush's API in June 2025; extensive and intensive Gen AI adoption measures as used in Liu, Huang, and Wang (2025). Used for Figure 1.1, panels a and b, and for Figure 4.12, panel b (country-level ChatGPT extensive/intensive usage microdata provided directly by the paper's authors). Data contains information from November 2022 to May 2025. Access date: June 2025 (Chapter 1); November 2025 (Chapter 4, provided by the paper's authors). File location: "Chapter 1\1 - Data\Raw"; "Chapter 4\1 - Data\Raw". File names: ChatGPT_extensive_intensive_202211_202505.dta. Liu, Yan, Jingyun Huang, and He Wang. 2025. "Who on Earth Is Using Generative AI? Global Trends and Shifts in 2025." Policy Research Working Paper 11231. World Bank. http://hdl.handle.net/10986/43859. License: CC BY 3.0 IGO.
    Access policy
    Data access requires purchase or human approval and is not included in the reproducibility package.
    License
    Proprietary (Semrush subscription terms)
    License URL
    https://www.semrush.com/projects/
    Data URL
    https://www.semrush.com/projects/
    Citation
    Semrush. (2025). "Generative AI Adoption Data" [dataset]. https://www.semrush.com/projects. Access date: June 2025.
    Name
    World Development Indicators
    Note
    Source: World Bank, World Development Indicators. Indicators used across chapters: Individuals using the Internet, % of population (IT.NET.USER.ZS; Chapter 1 Figure 1.1, accessed June 11, 2026; Chapter 7 Figure 7.1 panel b, accessed May 2026); Access to electricity, % of population (EG.ELC.ACCS.ZS; Chapter 7 Figure 7.1 panel a, accessed May 2026); GDP, current US$ (NY.GDP.MKTP.CD; Chapter 2 Figure 2.4, accessed July 16, 2026); GDP per capita, PPP, current international $ (NY.GDP.PCAP.PP.CD; Chapter 8 Figure B8.2.1, accessed 2026); Population, total (SP.POP.TOTL; Chapter 2 Figure 2.4, accessed July 16, 2026); GDP per capita, PPP, constant 2021 international $ (NY.GDP.PCAP.PP.KD; Chapter 4 Figure 4.12 panel a diagnostic check, accessed 2025); and API pulls of Employment to population ratio, 15+ (SL.EMP.TOTL.SP.ZS), Population ages 15-64 (SP.POP.1564.TO), and Population, total (SP.POP.TOTL) via the code in Figure_4_11b.do (Chapter 4 Figure 4.12 panel b, accessed June 2026; stored as processed files EMP_SHR.dta, WAP.dta, POP.dta). File location: "Chapter 1\1 - Data\Raw"; "Chapter 2\1 - Data"; "Chapter 7\1 - Data" (embedded in figure workbooks); "Chapter 8\1 - Data"; "Chapter 4\1 - Data"; "Chapter 4\1 - Data\Processed". File names: API_IT.NET.USER.ZS_DS2_en_csv_v2_293302_June_11_2026.csv; API_NY.GDP.MKTP.CD_DS2_en_csv_v2_234.csv; API_NY.GDP.PCAP.PP.CD_DS2_en_csv_v2_2869.csv; Figure 7.1 panel a.xlsx; Figure 7.1 panel b.xlsx; API_NY.GDP.PCAP.PP.KD_DS2_en_csv_v2_33608_July_15_2026.csv; EMP_SHR.dta; WAP.dta; POP.dta.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International License
    License URL
    https://data.worldbank.org/summary-terms-of-use
    Data URL
    https://data.worldbank.org/indicator
    Citation
    World Bank. 2026. World Development Indicators [dataset]. World Bank. https://data.worldbank.org. Accessed May-July 2026.
    Name
    World Bank Country and Lending Groups
    Note
    Source: World Bank country classification by income level. Versions used: FY2027 classification (CLASS_2026_07_01.xlsx; Chapter 1, Chapter 6; accessed July 1, 2026), FY2026 classification (CLASS_2025_07_02.xlsx; Chapter 2 Figure 2.4; accessed August 20, 2025), 2024 download based on 2023 GNI (CLASS.xlsx; Chapter 5 Figure 5.4), and historical classifications OGHIST based on 2025 GNI (OGHIST_2026_07_01.xlsx; Chapter 5, Chapter 8, Spotlight 7; accessed 2026); wb_income_fy26.csv (Spotlight 7). Chapter 4 includes two vintages: CLASS_2026_07_01.xlsx (FY2027; Figures 4.3b, 4.8, 4.10a, and 4.12a; accessed July 1, 2026) and CLASS_2025_10_07.xlsx (FY2026 vintage; accessed October 2025; likely the source of the income_full_fy26 variable in wageworker2020_24.dta used for the HIC/Non-HIC split in Figure 4.7). File location: "Chapter 1\1 - Data\Raw"; "Chapter 2\1 - Data"; "Chapter 5\1 - Data"; "Chapter 6\1 - Data\Raw"; "Chapter 8\1 - Data"; "Spotlight 7\1 - Data"; "Chapter 4\1 - Data\Raw". File names: CLASS_2026_07_01.xlsx; CLASS_2025_07_02.xlsx; CLASS.xlsx; OGHIST_2026_07_01.xlsx; wb_income_fy26.csv; CLASS_2025_10_07.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International License
    License URL
    https://data.worldbank.org/summary-terms-of-use
    Data URL
    https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups
    Citation
    World Bank. 2026. World Bank Country and Lending Groups [dataset]. World Bank. https://datahelpdesk.worldbank.org/knowledgebase/articles/906519. Accessed July 1, 2026.
    Name
    World Population Prospects
    Note
    Source: Total population by sex (both sexes), World Population Prospects (dashboard), Population Division, Department of Economic and Social Affairs, United Nations. Median values, years 2021-2025. Used for Chapter 1, Figure 1.1. Access date: January 13, 2026. File location: "Chapter 1\1 - Data\Raw". File names: Total_pop_model_based_est_Jan_13_2026.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 3.0 IGO
    License URL
    https://creativecommons.org/licenses/by/3.0/igo/
    Data URL
    https://population.un.org/wpp/
    Citation
    United Nations, Department of Economic and Social Affairs, Population Division. 2026. World Population Prospects [dataset]. United Nations. https://population.un.org/wpp/. Accessed January 13, 2026.
    Name
    GenAI Exposure Data
    Note
    Source: Underlying data provided directly by the authors of Gmyrek, Viollaz, and Winkler (2026), Policy Research Working Paper 11328 and background paper for WDR 2026. Used for Chapter 1 Figure 1.3 (panels a and b; Excel data provided by the authors in April 2026) and Chapter 5 Figure 5.3 (country-level exposure .rds files). The corresponding reproducibility package is at https://reproducibility.worldbank.org/catalog/449. Chapter 4 additionally uses two extracts provided directly by the paper's authors: scatter_WDR_data.xlsx (cross-country GenAI exposure scores; Figure 4.3 panel b; accessed April 2026) and emp_by_4dISCO_v2.xlsx (employment by 4-digit ISCO-08 code, sheets "HICs"/"nonHICs"; Figure 4.12 panel a; accessed July 2026). The paper is cited under two related identifiers: World Bank Policy Research Working Paper 11328 and ILO Working Paper 166 (https://doi.org/10.54394/00033147). The underlying individual-level LFS microdata used by Chapter 4 code is restricted and covered in a separate entry. Access date: April 2026 (Chapter 1); 2025 (Chapter 5). File location: "Chapter 1\1 - Data"; "Chapter 4\1 - Data\Raw". File names: Figure 1.3, panel a.xlsx; Figure 1.3, panel b.xlsx; scatter_WDR_data.xlsx; emp_by_4dISCO_v2.xlsx.
    Access policy
    Data access was granted directly to the study authors by the data owners. It was obtained with a custom data license that does not allow for redistribution and it is not included in the reproducibility package.
    Citation
    Gmyrek, Pawel; Viollaz, Mariana; Winkler, Hernan. 2026a. “Disruption without Dividend? How the Digital Divide and Task Differences Split GenAI’s Global Impact.” Policy Research Working Paper; 11328.
    Name
    Expertise and Automation by GenAI Data
    Note
    Source: Underlying data provided directly by the authors of Gmyrek, Jimenez, Winkler, and Yu (2026), background paper for World Development Report 2026: Artificial Intelligence. Used for Chapter 1, Figure 1.5. Access date: Provided by the authors in April 2026. File location: "Chapter 1\1 - Data". File names: Figure 1.5.xlsx.
    Access policy
    Data is included in the reproducibility package.
    License URL
    https://www.worldbank.org/ext/en/legal/terms-conditions
    Data URL
    https://www.worldbank.org/en/publication/wdr2026/brief/world-development-report-2026-background-papers
    Citation
    Gmyrek, Pawel, Hector Jimenez, Hernan Jorge Winkler, and Shu Yu. 2026. "Expertise and Automation by GenAI." Background paper for World Development Report 2026: Artificial Intelligence. World Bank. Accessed April 2026.
    Name
    Global Giants in the AI Supply Chain
    Note
    Source: Table taken from Frost, Rishabh, and Shreeti (2026), "Global Giants in the AI Supply Chain," BIS Bulletin 122 (February), Graph 1.A. Only publicly listed AI companies are kept. Used for Chapter 2, Table 2.1. Access date: 2026. File location: "Chapter 2\3 - Output". File names: Table 2.1.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    BIS Terms and conditions of use
    License URL
    https://www.bis.org/terms_conditions.htm
    Data URL
    https://www.bis.org/publ/bisbull122.pdf
    Citation
    Frost, Jon, Kumar Rishabh, and Vatsala Shreeti. 2026. "Global Giants in the AI Supply Chain." BIS Bulletin 122 (February). Bank for International Settlements. https://www.bis.org/publ/bisbull122.pdf.
    Name
    World Economic Outlook Database
    Note
    Source: International Monetary Fund (IMF), World Economic Outlook (WEO) Database, nominal GDP (current prices, U.S. dollars), accessed via IMF DataMapper. Hyperscaler capex projections for 2026 added through desktop research (links in workbook). Used for Chapter 2, Figure 2.1. Access date: May 2026. File location: "Chapter 2\1 - Data". File names: Figure 2.1.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    The Use of IMF Data
    License URL
    https://www.imf.org/en/About/copyright-and-terms
    Data URL
    https://www.imf.org/external/datamapper/NGDPD@WEO/OEMDC/ADVEC/WEOWORLD
    Citation
    International Monetary Fund. 2026. World Economic Outlook Database [dataset]. IMF. Accessed May 2026.
    Name
    Mineral Commodity Summaries 2026
    Note
    Source: U.S. Geological Survey (USGS), Mineral Commodity Summaries 2026. Chapters used: Gallium (p. 83), Rare Earths (p. 153), Tantalum (p. 187); 2025 production values are USGS estimates. Used for Chapter 2, Figure 2.3, panels a, b, and c. Access date: March 9, 2026. File location: "Chapter 2\1 - Data". File names: Figure 2.3, panel a.xlsx; Figure 2.3, panel b.xlsx; Figure 2.3, panel c.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Public domain (U.S. Government work)
    License URL
    https://www.usgs.gov/information-policies-and-instructions/copyrights-and-credits
    Data URL
    https://pubs.usgs.gov/periodicals/mcs2026/mcs2026.pdf
    Citation
    U.S. Geological Survey. 2026. Mineral Commodity Summaries 2026. USGS. https://pubs.usgs.gov/periodicals/mcs2026/mcs2026.pdf. Accessed March 9, 2026.
    Name
    GitHub Innovation Graph
    Note
    Source: GitHub Innovation Graph, topics dataset (AI-related GitHub pushes by country and quarter, Q1-Q4 2025). Used for Chapter 2, Figure 2.4. Access date: July 16, 2026. File location: "Chapter 2\1 - Data". File names: topics.csv.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons CC0 1.0 Universal
    License URL
    https://github.com/github/innovationgraph/blob/main/LICENSE.md
    Data URL
    https://github.com/github/innovationgraph/blob/main/data/topics.csv
    Citation
    GitHub. 2026. GitHub Innovation Graph [dataset]. GitHub. https://github.com/github/innovationgraph. Accessed July 16, 2026.
    Name
    The Artificial Intelligence Index Report 2026
    Note
    Source: Data folder for the Artificial Intelligence Index Report 2026 (Stanford HAI), section "1. Research and Development". Used for Chapter 2, Figures 2.2 (fig_1.8.6.csv) and 2.5 (fig_1.1.8.csv). Access date: April 2026. File location: "Chapter 2\1 - Data". File names: fig_1.8.6.csv; fig_1.1.8.csv.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Attribution-NoDerivatives 4.0 International
    License URL
    https://creativecommons.org/licenses/by-nd/4.0/?ref=chooser-v1
    Data URL
    https://hai.stanford.edu/ai-index/2026-ai-index-report, https://drive.google.com/drive/folders/1zJTOg0iR0j5SijCwFutwWvDt143lW277
    Citation
    Stanford Institute for Human-Centered Artificial Intelligence (HAI). 2026. The Artificial Intelligence Index Report 2026 [dataset]. Stanford University. https://hai.stanford.edu/ai-index/2026-ai-index-report. Accessed April 2026.
    Name
    Data on AI Models
    Note
    Source: Epoch AI, AI Models dataset (public table of AI models with country of developing organization and frontier-model flag). Used for Chapter 3, Figure 3.1, panels a and b. Access date: June 2, 2026. File location: "Chapter 3\1 - Data". File names: AI Models public table.csv.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International License
    License URL
    https://creativecommons.org/licenses/by/4.0/
    Data URL
    https://epoch.ai/data/ai-models
    Citation
    Epoch AI. 2026. "Data on AI Models" [dataset]. Epoch AI. https://epoch.ai/data/ai-models. Accessed June 2, 2026.
    Name
    Statistics of Common Crawl Monthly Archives
    Note
    Source: Common Crawl crawl statistics, distribution of URLs by primary language, crawl CC-MAIN-2026-08 (February 2026). Used for Chapter 3, Figure 3.2, panel a. Original download file: languages.csv, copied into the figure workbook (sheet "Raw"). Access date: February 2026. File location: "Chapter 3\1 - Data". File names: Figure 3.2 panel a.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Common Crawl Terms of Use
    License URL
    https://commoncrawl.org/terms-of-use
    Data URL
    https://commoncrawl.github.io/cc-crawl-statistics/plots/languages.html
    Citation
    Common Crawl. 2026. Statistics of Common Crawl Monthly Archives - Distribution of Languages [dataset]. Common Crawl. Accessed February 2026.
    Name
    Open Training Datasets by Language and Type
    Note
    Source: Hugging Face data on open training datasets by language (multilingual and monolingual types), accessed via the OECD.AI Policy Observatory visualization. Original download file: data.csv, copied into the figure workbook (sheet "Raw"). Used for Chapter 3, Figure 3.2, panel b. Access date: June 2026. File location: "Chapter 3\1 - Data". File names: Figure 3.2 panel b.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    OECD Terms & Conditions
    License URL
    https://www.oecd.org/en/about/terms-conditions.html
    Data URL
    https://oecd.ai/en/data?selectedArea=ai-models-and-datasets&selectedVisualization=open-training-datasets-by-language-and-type
    Citation
    OECD.AI Policy Observatory. 2026. Open Training Datasets by Language and Type (data from Hugging Face) [dataset]. OECD. Accessed June 2026.
    Name
    Lightcast Job Postings Data
    Note
    Source: Lightcast job postings data (proprietary, subscription-based). Restricted raw and firm-level pulls used across Chapter 4: (1) Raw global multi-country panel (100+ countries), used to build the country-level panels for Figure 4.7 (01_build_country_parquets.py); restricted to authorized users of the World Bank internal data environment. The pooled analysis panel derived from it (pooled_panel_target95_Qge_m10_with_emp20.parquet) likewise cannot be shared publicly. (2) Lightcast-derived MPS decile file, used as a task-exposure control in the preferred Figure 4.7 specification. File: mps_deciles_summary.csv. (3) SAR/India firm-level job postings, 2020Q1-2025Q3, used for Figure B4.3.1. File: job_postings_merged.dta. Accessed June 2025. (4) Origin and harmonized company names derived from Lightcast for India and other SAR countries, supporting the fuzzy-matching step in Figure B4.3.1. Files: comany_name_harmonized.csv; sar_name_clean.csv. All files are registered in the data hash report. File location: "Chapter 4\1 - Data\Figure_4_7\Processed"; "Chapter 4\1 - Data\Data_B4.3.1\raw\job_post"; "Chapter 4\1 - Data\Data_B4.3.1\raw\key". File names: mps_deciles_summary.csv; job_postings_merged.dta; comany_name_harmonized.csv; sar_name_clean.csv.
    Access policy
    Data access was granted directly to the study authors by the data owners. It was obtained with a custom data license that does not allow for redistribution and it is not included in the reproducibility package.
    License
    Proprietary (Lightcast)
    Data URL
    https://lightcast.io/products/data/overview
    Citation
    Lightcast. "Job Postings Data" [dataset]. https://lightcast.io/products/data/overview. Accessed 2025-2026.
    Name
    Lightcast Job Postings - Author-Aggregated Data
    Note
    Source: Authors' aggregations of proprietary Lightcast job postings data; only these aggregates are shared, the raw data cannot be redistributed. Two aggregates are used: (1) Gen AI vacancies and % of total vacancies by income group, 2021-2024, as aggregated in the World Bank Digital Progress and Trends Report 2025 (Figure 5.3, panel b). Used for Chapter 3, Figure 3.3. File: Figure 3.3.xlsx. Access date: 2025. (2) Share of postings requiring AI skills by country income group and distribution of AI-related postings by 4-digit occupation in LMICs. Used for Chapter 4, Figure 4.9, panels a and b. File: Figure 4.9 panels a and b.xlsx. Access/aggregation date: April 21, 2026. File location: "Chapter 3\1 - Data"; "Chapter 4\1 - Data". File names: Figure 3.3.xlsx; Figure 4.9 panels a and b.xlsx.
    Access policy
    Data is included in the reproducibility package
    License
    The Figure 3.3 aggregates are reproduced from a World Bank publication (CC BY 3.0 IGO); underlying Lightcast data proprietary
    Data URL
    https://lightcast.io/products/data/overview
    Citation
    Authors' compilation. 2026. "AI Job Postings Aggregates by Country Income Group" [dataset]. Based on Lightcast Job Postings Data (https://lightcast.io/products/data/overview) and the World Bank Digital Progress and Trends Report 2025. Accessed 2025-2026.
    Name
    AI Procurement Contracts
    Note
    Source: Authors' compilation. 2026. AI Procurement Contracts [dataset]. World Bank, World Development Report 2026 - Artificial Intelligence. A web scraper collected 6,127,564 procurement contracts from the World Bank, Open Contracting Partnership, and the EU Tenders Electronic Daily (TED) platform; DeepSeek and Gemini were used to classify contracts as AI-related, yielding 17,865 AI contracts with valid income group classification, further classified into tool categories. The cleaned dataset is in the process of being uploaded to the Development Data Hub. Intermediate/raw classification files (s3_tagged_notices.csv; NewSet.xlsx; per_stage_tag_results.csv; Reclassified_Tools_v4.csv) are WDR 2026 team outputs, not public elsewhere; exact reproduction of the LLM classification is not guaranteed. Used for Chapter 5, Figures 5.2 and 5.7. Access year: 2026. File location: "Chapter 5\1 - Data" and "Chapter 5\1 - Data\Data Organized by Figure\Procurement (Figures 5.2, 5.7)". File names: AI Procurement Contracts.xlsx; s3_tagged_notices.csv; NewSet.xlsx; per_stage_tag_results.csv; Reclassified_Tools_v4.csv.
    Access policy
    Data is forthcoming in the World Bank Development Data Hub.
    Data URL
    Forthcoming https://datacatalog.worldbank.org
    Citation
    Authors' compilation. 2026. AI Procurement Contracts [dataset]. World Bank, World Development Report 2026 - Artificial Intelligence. Based on procurement notices from the World Bank, Open Contracting Partnership, and EU Tenders Electronic Daily.
    Name
    Datahub Country List
    Note
    Source: Datahub core country-list dataset (country names with ISO 3166-1 alpha-2 codes), used to merge country identifiers when building the AI Procurement Contracts dataset. Used for Chapter 5, Figures 5.2 and 5.7. Access year: 2026. File location: "Chapter 5\1 - Data\Data Organized by Figure\Procurement (Figures 5.2, 5.7)\Files to Re-Create AI Procurement Contracts". File names: filtered_data.csv.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Open Data Commons Public Domain Dedication and License v1.0
    License URL
    https://opendatacommons.org/licenses/pddl/
    Data URL
    https://datahub.io/core/country-list
    Citation
    Datahub. 2026. Country List [dataset]. Datahub. https://datahub.io/core/country-list. Accessed 2026.
    Name
    International Survey on Revenue Administration
    Note
    Source: International Monetary Fund, ISORA latest published data, downloaded from the IMF data portal in late 2025. Used for Chapter 5, Figure 5.4. Access date: November 2025. File location: "Chapter 5\1 - Data" (duplicated in "Chapter 5\1 - Data\Data Organized by Figure\ISORA (Figure 5.4)"). File names: dataset_2025-08-25T16_24_03.921163402Z_DEFAULT_INTEGRATION_ISORA_ISORA_LATEST_DATA_PUB_4.0.0.csv.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    The Use of IMF Data
    License URL
    https://www.imf.org/en/About/copyright-and-terms
    Data URL
    https://data.imf.org/en/datasets/ISORA:ISORA_LATEST_DATA_PUB
    Citation
    International Monetary Fund. 2025. International Survey on Revenue Administration (ISORA) [dataset]. IMF. https://data.imf.org/en/datasets/ISORA:ISORA_LATEST_DATA_PUB. Accessed November 2025.
    Name
    Medical Imaging and Nuclear Medicine Global Resources Database
    Note
    Source: International Atomic Energy Agency (IAEA) IMAGINE database, radiology workforce availability across income groups. Used for Chapter 5, Figure 5.5, panels a and b. Access date: March 1, 2026. File location: "Chapter 5\1 - Data" (duplicated in "Chapter 5\1 - Data\Data Organized by Figure\Radiologists (Figure 5.5)"). File names: Radiologists.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    IAEA Terms of Use
    License URL
    https://www.iaea.org/about/terms-of-use
    Data URL
    https://www.iaea.org/resources/hhc/nuclear-medicine/databases/imagine/radiologists
    Citation
    International Atomic Energy Agency. 2026. IMAGINE - IAEA Medical Imaging and Nuclear Medicine Global Resources Database [dataset]. IAEA. Accessed March 1, 2026.
    Name
    AI and Data for Better Governance Survey
    Note
    Source: Data for Better Governance Team, World Bank. Anonymized results of the AI Data for Better Governance Survey as of June 17, 2026 (65 responses from 58 unique countries); internal World Bank survey, anonymized version in the process of being uploaded to the World Bank Microdata Library. Used for Overview, Chapter 5 (Figure 5.6), Chapter 8 (Figures 8.2, B8.1.1, 8.4, 8.5), and Chapter 9 (Figures 9.1, 9.5). Access year: 2026. File location: "0 - Overview\1 - Data"; "Chapter 5\1 - Data"; "Chapter 8\1 - Data"; "Chapter 9\1 - Data". File names: anonymized_agency_survey.xlsx.
    Access policy
    Data is forthcoming in the World Bank Microdata Library.
    Data URL
    Forthcoming at https://microdata.worldbank.org/
    Citation
    World Bank, Data for Better Governance Team. 2026. AI and Data for Better Governance Survey [dataset]. World Bank Microdata Library (forthcoming).
    Name
    EdTech Meta-Analysis Data
    Note
    Source: Meta-analysis of EdTech in education interventions from Burneo, Dinarte-Diaz, Lopez, and Molina. (2026), background paper for WDR 2026. One row per effect size estimate with standardized effect size, standard error, confidence intervals, intervention description, and study information. Used for Chapter 5, Figure B5.2.1. Access year: 2026. File location: "Chapter 5\1 - Data" (duplicated in "Chapter 5\1 - Data\Data Organized by Figure\Tutoring Box (Figure B5.2.1)" and "Chapter 5\1 - Data\Figure_B5.2.1\data"). File names: meta_results.csv.
    Access policy
    Data is included in the reproducibility package.
    Data URL
    https://www.worldbank.org/en/publication/wdr2026/brief/world-development-report-2026-background-papers
    Citation
    Burneo, Dinarte-Diaz, Lopez, and Molina. World Bank. 2026. EdTech Meta-Analysis Data [dataset]. Accessed 2026.
    Name
    BACI: International Trade Database at the Product-Level
    Note
    Source: CEPII, BACI international trade database, HS 2022 version, year 2024, with the accompanying country code crosswalk. Used for Chapter 6, Figures 6.1 and 6.2. Access date: March 2026. File location: "Chapter 6\1 - Data\Raw". File names: BACI_HS22_Y2024_V202601.csv; country_codes_V202601.csv.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Etalab 2.0
    License URL
    https://www.etalab.gouv.fr/wp-content/uploads/2018/11/open-licence.pdf
    Data URL
    https://www.cepii.fr/DATA_DOWNLOAD/baci/doc/baci_webpage.html
    Citation
    CEPII. 2026. BACI: International Trade Database at the Product-Level, HS22 version, 2024 [dataset]. CEPII. Accessed March 2026.
    Name
    Balanced Trade in Services dataset
    Note
    Source: WTO-OECD Balanced Trade in Services dataset; sectors "Financial services" and "Insurance and pension services", year 2024. Used for Chapter 6, Figure 6.2. Access date: March 2026. File location: "Chapter 6\1 - Data\Raw". File names: WtoData_20260317180803.csv.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    WTO Terms of Use
    License URL
    https://www.wto.org/english/info_e/copyrights_permissions_e.htm
    Data URL
    https://www.wto.org/english/res_e/statis_e/gstdh_batis_e.htm
    Citation
    World Trade Organization and OECD. 2026. Balanced Trade in Services (BaTIS) [dataset]. Accessed March 2026.
    Name
    Federal Lobbying by Issue
    Note
    Source: OpenSecrets, Federal Lobbying by Issue, datasets for all top issues downloaded one by one for each year from 1998 to 2025. Used for Chapter 6, Figure 6.3. Access date: 2026. File location: "Chapter 6\1 - Data\Raw". File names: lobbying_top_issues_1998.csv through lobbying_top_issues_2025.csv (28 files).
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution-Noncommercial-Share Alike 3.0
    License URL
    https://creativecommons.org/licenses/by-nc-sa/3.0/deed.en
    Data URL
    https://www.opensecrets.org/federal-lobbying/top-issues
    Citation
    OpenSecrets. 2026. Federal Lobbying by Issue [dataset]. OpenSecrets. https://www.opensecrets.org/federal-lobbying/top-issues. Accessed 2026.
    Name
    Pew Research Center Global Attitudes Survey, Spring 2025
    Note
    Source: Poushter et al. 2025, Pew Research Center Global Attitudes Survey, Spring 2025. Question 31 on feelings about increased use of AI in daily life. Used for Chapter 6, Figure 6.4. Access date: July 2026. File location: "Chapter 6\1 - Data\Raw". File names: Pew Research Center Global Attitudes Spring 2025 Dataset CSV.csv. The data can be downloaded directly from the link below after logging in; after that, the user should place the file in the appropriate folder to generate Figure 6.4.
    Access policy
    Data is publicly available but does not allow redistribution and is not included in the reproducibility package.
    License
    Pew Research Center terms of use
    License URL
    https://www.pewresearch.org/about/terms-and-conditions/
    Data URL
    https://www.pewresearch.org/dataset/spring-2025-survey-data/
    Citation
    Pew Research Center. 2025. Global Attitudes Survey, Spring 2025 [dataset]. https://www.pewresearch.org/dataset/spring-2025-survey-data/. Accessed July 2026.
    Name
    Pew Research Center American Trends Panel Wave 152
    Note
    Source: Pew Research Center, American Trends Panel Wave 152 (latest AI wave available as of July 2026), used to add U.S. responses to the Spring 2025 Global Attitudes analysis. Used for Chapter 6, Figure 6.4. Access date: July 2026. File location: "Chapter 6\1 - Data\Raw". File names: ATP W152.csv. The data can be downloaded directly from the link below after logging in; after that, the user should place the file in the appropriate folder to generate Figure 6.4.
    Access policy
    Data is publicly available but does not allow redistribution and is not included in the reproducibility package.
    License
    Pew Research Center terms of use
    License URL
    https://www.pewresearch.org/about/terms-and-conditions/
    Data URL
    https://www.pewresearch.org/dataset/american-trends-panel-wave-152/
    Citation
    Pew Research Center. 2025. American Trends Panel Wave 152 [dataset]. Pew Research Center. Accessed July 2026.
    Name
    Global Findex Database 2024
    Note
    Source: World Bank Global Findex Database, 2024 values, accessed via Data360: Mobile phone ownership, 15+ (https://data360.worldbank.org/en/int/indicator/WB_FINDEX_CON1?SEX=_T&AGE=Y_GE15&URBANISATION=_T&COMP_BREAKDOWN_1=_T&COMP_BREAKDOWN_2=_T&COMP_BREAKDOWN_3=_T&view=datatable&average=incomeGroup) and Main mobile phone is a smartphone (https://data360.worldbank.org/en/int/indicator/WB_FINDEX_CON9A?view=datatable&minYear=2024&maxYear=2024&SEX=_T&AGE=Y_GE15&URBANISATION=_T&COMP_BREAKDOWN_1=_T&COMP_BREAKDOWN_2=_T&COMP_BREAKDOWN_3=_T&average=incomeGroup), income group aggregates. Used for Chapter 7, Figure 7.2, panel a. Access date: July 2026. File location: "Chapter 7\1 - Data". File names: Figure 7.2 panel a.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International License
    License URL
    https://data.worldbank.org/summary-terms-of-use
    Citation
    World Bank. 2024. Global Findex Database [dataset]. World Bank Data360. https://data360.worldbank.org/en/int/indicator Accessed July 2026.
    Name
    Global Digital Inclusion 2024 Device Pricing Data
    Note
    Source: Global Digital Inclusion Partnership (GDIP). 2024. "The Price of the World in Your Pocket: A Price Only Some Can Afford: 2024 Device Pricing." Excel file "Device pricing data (Excel)". Used for Chapter 7, Figure 7.2, panel b. Access date: July 2026. File location: "Chapter 7\1 - Data". File names: Figure 7.2 panel b.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    Data URL
    https://globaldigitalinclusion.org/2024/09/10/the-price-of-the-world-in-your-pocket-a-price-only-some-can-afford-2024/
    Citation
    Global Digital Inclusion Partnership. 2024. 2024 Device Pricing [dataset]. GDIP. Accessed July 2026.
    Name
    OECD Survey of Adult Skills 2023
    Note
    Source: OECD (2024), "Do Adults Have the Skills They Need to Thrive in a Changing World?: Survey of Adult Skills 2023," OECD Skills Studies. Country-level data on percentage of 35-44-year-olds by training sponsor type. Used for Chapter 7, Figure 7.4. Access date: May 27, 2026. File location: "Chapter 7\1 - Data". File names: Figure 7.4.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    OECD Terms & Conditions
    License URL
    https://www.oecd.org/en/about/terms-conditions.html
    Data URL
    https://stat.link/n4cktb
    Citation
    OECD. 2024. Do Adults Have the Skills They Need to Thrive in a Changing World?: Survey of Adult Skills 2023 [dataset]. OECD Skills Studies, OECD Publishing. Accessed May 27, 2026.
    Name
    Data Centers by Population
    Note
    Source: Combined dataset obtained directly from the authors of Straub et al. (2026), Infrastructure Foundations (World Bank, Sustainable Infrastructure Series). https://doi.org/10.60572/PY8T-MB68. Number of data centers from TeleGeography (proprietary, access year 2026); The map was created using the file Map 7.1.csv, which was created by aggregating the number of data centers by country using restricted data from the report and population from World Bank WDI, SP.POP.TOTL (publicly available, access year 2025). Map updated by the cartography team. Used for Chapter 7, Map 7.1. Aggregated data is included in the package. File location: "Chapter 7\1 - Data". File names: Map 7.1.csv.
    Access policy
    Data access was granted directly to the study authors by the data owners/managers. It was obtained with a custom data license that does not allow for redistribution, and it is not included in the reproducibility package. Aggregated intermediate data is included in the package.
    Data URL
    https://www2.telegeography.com/
    Citation
    Telegeography. (2026). Data Centers by Population [dataset]. Unpublished data.
    Name
    Teaching and Learning International Survey
    Note
    Source: OECD Teaching and Learning International Survey (TALIS) 2024, teacher questionnaire (ttgintt4.dta). Used for Chapter 7, Table 7.1 (items TT4G36, TT4G24G, TT4G21G) and Spotlight 3, Figures S3.1A and S3.1B (items TT4G35G, TT4G35H). Access date: January 12, 2026. File location: "Chapter 7\1 - Data"; "Spotlight 3\1 - Data". File names: ttgintt4.dta.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    OECD Terms & Conditions
    License URL
    https://www.oecd.org/en/about/terms-conditions.html
    Data URL
    https://www.oecd.org/en/data/datasets/talis-2024-database.html
    Citation
    OECD. 2024. Teaching and Learning International Survey (TALIS) 2024 [dataset]. OECD. https://www.oecd.org/en/data/datasets/talis-2024-database.html. Accessed January 12, 2026.
    Name
    World Bank GovTech Maturity Index (GTMI)
    Note
    Source: World Bank GovTech Maturity Index database, December 2025 version. Used for Chapter 8, Figure 8.3. Access year: 2026. File location: "Chapter 8\1 - Data" (duplicated in "Chapter 8\1 - Data\Data Organized by Figure\GTMI Database (Figure 8.3)"). File names: WBG_GovTech_Dataset_Dec2025.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International License
    License URL
    https://data.worldbank.org/summary-terms-of-use
    Data URL
    https://datacatalog.worldbank.org/search/dataset/0037889/govtech-dataset
    Citation
    World Bank. 2025. GovTech Dataset (GovTech Maturity Index) [dataset]. World Bank Data Catalog. Accessed 2026.
    Name
    AI-Mediated Access to Government Services Data
    Note
    Data was obtained from a forthcoming background paper. Jordan, Luke, Manuel Ramos-Maqueda, Tiago C. Peixoto, and Tapan Parikh. "RADAR: Benchmarking AI-Mediated Access to Government Services." Background paper for WDR 2026. Ratings from four major LLMs across services in navigating online public service queries. Used for Chapter 8, Figure B8.2.1. Access year: 2026. File location: "Chapter 8\1 - Data". File names: scores.csv.
    Access policy
    Data is included in the reproducibility package.
    Data URL
    https://www.worldbank.org/en/publication/wdr2026/brief/world-development-report-2026-background-papers
    Citation
    Jordan, Luke, Manuel Ramos-Maqueda, Tiago C. Peixoto, and Tapan Parikh. Forthcoming. "RADAR: Benchmarking AI-Mediated Access to Government Services." Background paper for World Development Report 2026: Artificial Intelligence. World Bank.
    Name
    GRIDMAP Data Markets Dashboard
    Note
    Source: Global Regulations, Institutional Development, and Market Authorities Perspective Toolkit (GRIDMAP): Data Markets Module, World Bank. Figure 9.3 uses the gap to Minimum Package (MP) by income group from the dashboard Overview plus MP values from the accompanying World Bank document; Figure 9.4 uses Survey responses (Pillar 2: Institutional Oversight, Category 6: Resources, Data Protection Authority budget question). Used for Chapter 9, Figures 9.3 and 9.4. Access date: July 1, 2026. File location: "Chapter 9\1 - Data". File names: figure_9.3.xlsx; figure_9.4.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International License
    License URL
    https://data.worldbank.org/summary-terms-of-use
    Data URL
    https://www.worldbank.org/en/data/interactive/2025/10/07/gridmap-data-markets-dashboard
    Citation
    World Bank. 2025. GRIDMAP Data Markets Dashboard [dataset]. World Bank. https://www.worldbank.org/en/data/interactive/2025/10/07/gridmap-data-markets-dashboard. Accessed July 1, 2026.
    Name
    United Nations Governing AI for Humanity Report Data
    Note
    Source: United Nations. 2024. Governing AI for Humanity: Final Report. Data taken directly from Figure (a) (representation in seven non-UN international AI governance initiatives) and Figure 12 (Top 60 AI countries, 2023 Tortoise Index, in major plurilateral AI governance initiatives). Used for Chapter 9, Figure 9.6. Access date: July 1, 2026. File location: "Chapter 9\1 - Data". File names: figure_9.6.xlsx.
    Access policy
    Data is publicly available and included in the reproducibility package.
    Data URL
    https://www.un.org/sites/un2.un.org/files/governing_ai_for_humanity_final_report_en.pdf
    Citation
    United Nations. 2024. Governing AI for Humanity: Final Report. United Nations. Accessed July 1, 2026.
    Name
    Energy and AI Dataset
    Note
    Source: IEA (International Energy Agency). 2025. Energy and AI. Data annex, "Regional Data" spreadsheet (Annex A, Table A.4: Data centres electricity consumption by region). Used for Spotlight 4, Figure S4.1. Access date: April 2026. File location: "Spotlight 4\1 - Data" (revised copy with figure in "Spotlight 4\3 - Output"). File names: Data_annex_Energy_and_AI.xlsx. The data can be downloaded directly from the link below after logging in; after that, the user should place the file in the appropriate folder to generate Spotlight 4, Figure S4.1. The final numbers used to create the figure are included in the revised copy with the figure, only with the relevant numbers.
    Access policy
    Data is publicly available but does not allow redistribution and is not included in the reproducibility package.
    License
    Terms of Use for Non-CC Material
    License URL
    https://www.iea.org/terms/terms-of-use-for-non-cc-material
    Data URL
    https://www.iea.org/data-and-statistics/data-product/energy-and-ai
    Citation
    International Energy Agency. 2025. Energy and AI Dataset [dataset]. IEA. https://www.iea.org/data-and-statistics/data-product/energy-and-ai. Accessed April 2026.
    Name
    ASPIRE: The Atlas of Social Protection - Indicators of Resilience and Equity
    Note
    Source: World Bank ASPIRE database, Social Insurance Coverage (% of total population), indicator PER_SI_ALLSI.COV_POP_TOT. Used for Spotlight 5, Figure S5.1. Access date: May 12, 2026. File location: "Spotlight 5\1 - Data" File names: API_PER_SI_ALLSI.COV_POP_TOT_DS2_en_excel_v2_2323.xls.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International License
    License URL
    https://data.worldbank.org/summary-terms-of-use
    Data URL
    https://data.worldbank.org/indicator/per_si_allsi.cov_pop_tot
    Citation
    World Bank. 2026. ASPIRE: The Atlas of Social Protection Indicators of Resilience and Equity [dataset]. World Bank. https://data.worldbank.org/indicator/per_si_allsi.cov_pop_tot. Accessed May 12, 2026.
    Name
    AI Agencies Institutional Frameworks Dataset
    Note
    Source: Authors' compilation. 2026. AI Agencies Institutional Frameworks Dataset [dataset]. World Bank, World Development Report 2026 - Artificial Intelligence. Based on systematic desktop research across 218 World Bank-recognized countries and territories to identify dedicated AI agencies, using various government agency websites (links documented in the "Links" tab of the workbook). Used for Spotlight 7, Figure S7.2. Access year: 2026. File location: "Spotlight 7\1 - Data". File names: Institutional_Framework.xlsx.
    Access policy
    Data is included in the reproducibility package.
    Citation
    Authors' compilation. 2026. AI Agencies Institutional Frameworks Dataset [dataset]. World Bank, World Development Report 2026 - Artificial Intelligence. Based on various government agency websites.
    Name
    World Bank Enterprise Surveys - AI Follow-up and Consolidated Data
    Note
    Source: World Bank Enterprise Surveys (WBES) Data Portal. Comprises (1) the 2026 AI Follow-up surveys for 10 economies (India, Jordan, Kenya, Malaysia, Mexico, Nigeria, Thailand, Turkiye, United States, plus the New Comprehensive file), listed under "Other surveys" conducted in 2026, and (2) consolidated WBES datasets from the "Microdata across Economies" tab (ES Indicators Database - Global Methodology; Firm Level Factor Ratios). Used across chapters for Figures O.2, 1.2, 1.4, 4.2 (a, b), 4.4, 4.5, 4.13 (a, b), and 7.3. Access date: June 10, 2026. File location: "WBES AI Replication Package\source". File names: India-2026-AI follow-up data.dta; Jordan-2026-AI follow-up data.dta; Kenya-2026-AI follow-up data.dta; Malaysia-2026-AI follow-up data.dta; Mexico-2026-AI follow-up data.dta; Nigeria-2026-AI follow-up data.dta; Thailand-2026-AI follow-up data.dta; Turkiye-2026-AI follow-up data.dta; United States-2026-AI follow-up data.dta; New_Comprehensive_June_4_2026.dta; ES-Indicators-Database-Global-Methodology_June_4_2026.dta; Firm Level Factor Ratios_June_4_2026.dta.
    Access policy
    Data is publicly available but does not allow redistribution and is not included in the reproducibility package.
    License
    World Bank Enterprise Surveys Data Access Protocol
    License URL
    https://login.enterprisesurveys.org/en/terms
    Data URL
    https://login.enterprisesurveys.org/en/signin
    Citation
    World Bank. 2026. World Bank Enterprise Surveys - AI Follow-up Surveys [dataset]. World Bank Enterprise Surveys Data Portal. https://www.enterprisesurveys.org. Accessed June 10, 2026.
    Name
    ILOSTAT Employment Indicators
    Note
    Three indicator pulls: (1) Employment by Sex and Status in Employment (EMP_2EMP_SEX_STE_NB_A), Figure 4.3 panel b, file EMP_2EMP_SEX_STE_NB_A-full-2026-06-06.dta, downloaded via https://rplumber.ilo.org/data/indicator?id=EMP_2EMP_SEX_STE_NB_A, accessed June 2026; (2) general 'Employment Indicators' used in Figure 4.6, accessed 2026, specific indicator code not stated; (3) Employment by Sex and Economic Activity (EMP_TEMP_SEX_EC2_NB_A), Figure 4.8, file EMP_TEMP_SEX_EC2_NB_A.csv, via ILOSTAT bulk download facility, accessed March 2026. File location: "Chapter 4\1 - Data\Raw".
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    ILO Open Access Policy
    License URL
    https://www.ilo.org/rights-and-permissions
    Data URL
    https://ilostat.ilo.org/data/
    Citation
    International Labour Organization (ILO). 2026. "ILOSTAT Database" [dataset]. https://ilostat.ilo.org/data/. Accessed March-June 2026 (see note).
    Name
    Anthropic Economic Index
    Note
    Anthropic Economic Index, September 2025 Release, accessed via Hugging Face. Used in Figure 4.6 panels a and b, combined into AI productivity gain estimates Figure 4.6.xlsx. File location: "Chapter 4\1 - Data".
    Access policy
    Data is publicly available and included in the reproducibility package.
    License URL
    https://huggingface.co/terms-of-service
    Data URL
    https://huggingface.co/datasets/Anthropic/EconomicIndex
    Citation
    Anthropic. 2025. "Anthropic Economic Index" [dataset]. September 2025 Release. https://huggingface.co/datasets/Anthropic/EconomicIndex. Accessed 2025.
    Name
    GPTs are GPTs AI Exposure Scores
    Note
    Eloundou et al. (2024) LLM occupational exposure scores. The paper did not publish a ready-made results file; scores were reconstructed by re-running the authors' published code against their published inputs from GitHub. Used in: (1) Figure 4.6 as eloundou2024_score.dta, accessed 2026; (2) Figure B4.2.1 as the human_beta and gpt4_beta measures, sourced from GPTs_are_GPTs_paper_AI_exposure_data.csv. File location: "Chapter 4\1 - Data".
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    MIT License
    License URL
    https://github.com/openai/GPTs-are-GPTs/blob/main/LICENSE
    Data URL
    https://github.com/openai/GPTs-are-GPTs
    Citation
    Eloundou, T., Manning, S., Mishkin, P., and Rock, D. 2024. "GPTs are GPTs: Labor Market Impact Potential of LLMs." Science 384(6702): 1306-1308. https://doi.org/10.1126/science.adj0998. Data reconstructed by authors from https://github.com/openai/GPTs-are-GPTs. Accessed 2026.
    Name
    Exposure to Generative AI
    Note
    Language Modeling and Image Generation AIOE/AIIE indices, sourced from the AIOE-Data GitHub repo. Used in Figure B4.2.1 as felten2023_lm and felten2023_img. Files: Language Modeling AIOE and AIIE.xlsx (sheet 'LM AIOE'), Image Generation AIOE and AIIE.xlsx (sheet 'IG AIOE'). File location: "Chapter 4\1 - Data".
    Access policy
    Data is publicly available and included in the reproducibility package.
    Data URL
    https://github.com/AIOE-Data/AIOE
    Citation
    Felten, E. W., Raj, M., and Seamans, R. 2023. "Occupational Heterogeneity in Exposure to Generative AI." SSRN Working Paper 4414065. https://doi.org/10.2139/ssrn.4414065. Accessed 2026.
    Name
    Occupational, Industry, and Geographic Exposure to AI
    Note
    Occupation-level AI exposure index used in Figure B4.3.1 (ai_exposure.dta). File location: "Chapter 4\1 - Data\Data_B4.3.1\raw\ai_exp".
    Access policy
    Data is publicly available and included in the reproducibility package.
    Citation
    Felten, E., Raj, M., and Seamans, R. 2021. "Occupational, Industry, and Geographic Exposure to Artificial Intelligence: A Novel Dataset and Its Potential Uses." Strategic Management Journal 42(12): 2195-2217. Accessed November 2025.
    Name
    ILO Generative AI and Jobs
    Note
    Occupational exposure scores, ILO Working Paper 96. Not obtained directly from ILO; sourced from the reproducibility package of a separate World Bank paper (catalog https://reproducibility.worldbank.org/catalog/168, path RR_LAC_2024_168-v01/.../bases/ILO/Gmyrek_Berg_Bescond_scores_2023.csv). Used in Figure B4.2.1 as gbb2023_score. File location: "Chapter 4\1 - Data".
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Modified BSD3
    License URL
    https://opensource.org/license/bsd-3-clause/
    Data URL
    https://reproducibility.worldbank.org/catalog/168
    Citation
    Gmyrek, P., Winkler, H., & Garganta, S. 2024. "Reproducibility package for Buffer or Bottleneck? Generative AI, employment exposure and the digital divide in Latin America." World Bank. https://doi.org/10.60572/106F-PK97. Accessed 2026.
    Name
    ILO Refined Global Index of Occupational Exposure
    Note
    Used in: (1) Figure B4.2.1 as ilo25, manually extracted from the paper's Annex Table A1 ('ISCO-08 occupations by exposure gradient') since it is a PDF-embedded table with no downloadable file; (2) Figure 4.12 panel a as the basis for scenarios.xlsx tabs 'scenario_2.5%' and 'scenario_10%' (columns A-I), accessed April 2026 - columns J-L in that file are the WDR team's own scenario indicators calculated in Excel, not original ILO data. File location: "Chapter 4\1 - Data".
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    ILO Open Access Policy
    License URL
    https://www.ilo.org/rights-and-permissions
    Data URL
    https://www.ilo.org/sites/default/files/2025-05/WP140_web.pdf
    Citation
    Gmyrek, P., Berg, J., Kaminski, K., Konopczynski, F., Ladna, A., Nafradi, B., Roslaniec, K., and Troszynski, M. 2025. "Generative AI and Jobs: A Refined Global Index of Occupational Exposure." ILO Working Paper 140. https://www.ilo.org/publications/generative-ai-and-jobs-refined-global-index-occupational-exposure. Accessed April 2026.
    Name
    Labour Force Survey Microdata GenAI Exposure
    Note
    Individual-level labor force survey microdata provided by Gmyrek et al. (2026a). Two variants: All_countries_individuals.rds (Figure 4.12a, via WDR_emp_estimations_review.R) and All_countries_individuals_w_raw_sector.rds (Figure 4.8, via 01_lfs_data_collapse_country_isic.R, produces exposure_data_country_yr_isic2d.csv). Three further processed files derived from this microdata (exposure_2d.dta, exposure_2d_by_income.dta, exposure_2d_by_lmic.dta) are also excluded from the package per the Figure 4.8 README.
    Access policy
    Data access was granted directly to the study authors by the data owners. It was obtained with a custom data license that does not allow for redistribution and it is not included in the reproducibility package.
    License
    Custom License
    Citation
    Gmyrek, Pawel; Viollaz, Mariana; Winkler, Hernan. 2026. "Disruption without Dividend?" [dataset]. Unpublished data. Accessed March 2026.
    Name
    Occupational Classification Crosswalk
    Note
    US Bureau of Labor Statistics 2010 SOC-to-ISCO-08 crosswalk (sheet '2010 SOC to ISCO-08'), used in Figure B4.2.1 to align Felten (2023) and Eloundou (2024) measures (natively SOC 2010) to ISCO-08. File: isco_soc_crosswalk.xls. File location: "Chapter 4\1 - Data". (Derived Stata file crosswalk_soc2010_to_isco08.dta is a processed output and does not need its own entry.)
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    U.S. Bureau of Labor Statistics Terms of Service
    License URL
    https://www.bls.gov/developers/termsOfService.htm
    Data URL
    https://www.bls.gov/soc/isco_soc_crosswalk.xls
    Citation
    U.S. Bureau of Labor Statistics. "2010 SOC to ISCO-08 Crosswalk" [dataset]. https://www.bls.gov/soc/isco_soc_crosswalk.xls. Accessed 2026.
    Name
    ChatGPT Users per Internet User
    Note
    Values on ChatGPT users per internet user by income group, taken directly from the paper's published figures/tables (public, CC BY 3.0 IGO). Used in Figure 4.3 panel a, file Figure 4.3 panel a.xlsx. File location: "Chapter 4\1 - Data". The underlying raw series was originally pulled by the paper's authors via Semrush's API (proprietary), but only the paper's already-published aggregate values are reused here.
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 3.0 IGO (CC BY 3.0 IGO)
    License URL
    https://creativecommons.org/licenses/by/3.0/igo/
    Data URL
    http://hdl.handle.net/10986/43859
    Citation
    Liu, Yan; Huang, Jingyun; Wang, He. 2025. "Who on Earth Is Using Generative AI? Global Trends and Shifts in 2025." Policy Research Working Paper 11231. World Bank. http://hdl.handle.net/10986/43859. Accessed March 2026.
    Name
    FactSet Firm-to-Firm International Relationship Data
    Note
    Firm-to-firm international relationship data for SAR countries, 2020Q1-2025Q3, used to identify GVC-supplier status of firms in the Lightcast job postings data. Files: IND_firm_International_AIExposure.dta and SAR_firm_International_AIExposure.dta. File location: "Chapter 4\1 - Data\Data_B4.3.1\raw\factset". Proprietary; not included.
    Access policy
    Data access was granted directly to the study authors by the data owners. It was obtained with a custom data license that does not allow for redistribution and it is not included in the reproducibility package.
    License URL
    https://www.factset.com/legal
    Data URL
    https://www.factset.com/download
    Citation
    FactSet. "Firm-to-Firm International Relationship Data" [dataset]. https://www.factset.com/download. Accessed November 2025.
    Name
    LLM-Classified Local and Multi-national Company Data
    Note
    Firm headquarters/ownership classification (local vs. MNC affiliate) for firms in the Lightcast job postings data, generated via an LLM pipeline (Gemini 2.5 Flash-Lite, queried by firm name) identifying country of incorporation, primary country of operations, listing status, and foreign-ownership indicators. Files: IND_local_FULL_DETAILED_*.csv (14 dated files, Nov-Dec 2025) and SA_local_FULL_DETAILED_2026-01-29.csv. File location: "Chapter 4\1 - Data\Data_B4.3.1\raw\mnc\ind"; "Chapter 4\1 - Data\Data_B4.3.1\raw\mnc\sar". Proprietary (built from Lightcast firm names); not included.
    Access policy
    Data access was granted directly to the study authors by the data owners. It was obtained with a custom data license that does not allow for redistribution and it is not included in the reproducibility package.
    License
    Custom License
    Citation
    Authors' compilation. 2025-2026. "LLM-Classified Local and Multi-national Company Data" [dataset]. Unpublished data. Based on Lightcast job postings data and Gemini 2.5 Flash-Lite LLM classification. Accessed November 2025-January 2026.
    Name
    Fuzzy-Matched Lightcast-FactSet Linking Data
    Note
    Fuzzy-matching results linking Lightcast firms to FactSet firms (MinHash candidate identification + string similarity + manual verification), for India and other SAR countries. Files: Fuzzy_merge_v2.xlsx and Fuzzy_merge_SAR.csv. File location: "Chapter 4\1 - Data\Data_B4.3.1\raw\key". Proprietary (built from two proprietary sources); not included.
    Access policy
    Data access was granted directly to the study authors by the data owners. It was obtained with a custom data license that does not allow for redistribution and it is not included in the reproducibility package.
    License
    Custom License
    Citation
    Authors' compilation. 2025-2026. "Lightcast-FactSet Fuzzy-Matched Firm Linkage Dataset" [dataset]. Unpublished data. Based on Lightcast job postings data and FactSet firm relationship data.
    Name
    World Bank Digital Business Database
    Note
    World Bank 2024 Digital Business Database (Zhu et al. 2022), based on Crunchbase, CB Insights, Pitchbook, and Bitter Bridge, covering approximately 3,600 AI-related businesses in LMICs. File: digibiz_description.csv. Used in Figure 4.10 panel b. The classified output ('Category' column only) has separately been uploaded to the World Bank Development Data Hub as 'AI startup classification in developing countries' https://datacatalog.worldbank.org/int/data/dataset/0067142/ai_startup_classification_in_developing_countries
    Access policy
    Data access was granted directly to the study authors by the data owners. It was obtained with a custom data license that does not allow for redistribution and it is not included in the reproducibility package.
    License
    Custom License
    Citation
    Zhu, T. et al. 2022. "World Bank Digital Business Database." Based on Crunchbase, CB Insights, Pitchbook, and Bitter Bridge. Unpublished data. Accessed 2025.
    Name
    Potential Growth Estimates
    Note
    Potential growth estimates from Box 1.1 of Global Economic Prospects, June 2026. Values copied into AI productivity gain estimates Figure 4.6.xlsx ('Raw>>potential_growth'). Used in Figure 4.6 panel b. File location: "Chapter 4\1 - Data".
    Access policy
    Data is publicly available and included in the reproducibility package.
    License
    Creative Commons Attribution 4.0 International (CC BY 4.0)
    License URL
    https://www.worldbank.org/en/about/legal/terms-of-use-for-datasets
    Data URL
    https://thedocs.worldbank.org/en/doc/2b672b3b0415d6b66c45b66579db4ef5-0050012026/related/GEP-Jun-2026-Box-1-1.pdf
    Citation
    Kose, M. Ayhan, and Kersten Stamm. 2026. "Box 1.1: How Much Will AI Affect Global Growth?" In Global Economic Prospects, June 2026, 8-13. World Bank. https://thedocs.worldbank.org/en/doc/2b672b3b0415d6b66c45b66579db4ef5-0050012026/related/GEP-Jun-2026-Box-1-1.pdf. Accessed 2026.
    Name
    UN Comtrade Database - AI-Related Trade Data
    Note
    Trade-flow data (exports/imports, 2024, all reporters/partners) queried using three sets of AI-related HS Commodity Codes, each identified from a separate published report: (1) wto_2024.csv - HS codes from WTO, World Trade Report 2025 'Making Trade and AI Work Together to the Benefit of All' (AI-enabling products), accessed July 7, 2026; (2) oecd_2024.csv - HS codes from OECD (2025) 'Mapping the Semiconductor Value Chain' Annex D, accessed July 7, 2026; (3) fed_2024.csv - HS codes from FED (2026) 'The Global Trade Effects of the AI Infrastructure Boom' Table 1, accessed July 7, 2026. Query settings held constant: Goods, Annual, HS classification, 2024 period, all reporters/partners, World as 2nd partner, Exports and Imports flows. Used in Figure 4.10 panel a. UN Comtrade's terms do not permit redistribution of bulk/raw extracted records, so these files have been removed from the reproducibility package.
    Access policy
    Data is publicly available but does not allow redistribution, and it is not included in the reproducibility package.
    License
    UN Comtrade License Agreement
    License URL
    https://comtradeplus.un.org/LicenseAgreement
    Data URL
    https://comtradeplus.un.org/
    Citation
    United Nations. 2026. "UN Comtrade Database" [dataset]. https://comtradeplus.un.org/. HS codes per World Trade Organization, World Trade Report 2025; OECD, Mapping the Semiconductor Value Chain (2025); Federal Reserve, FEDS Notes (2026). Accessed July 7, 2026.
    Name
    Occupation and Income Classification Data
    Note
    Authors' estimates combining three shareable auxiliary inputs to Figure 4.7 (confirmed as pre-built inputs, not generated by any script in this package): (1) occupation_scores.dta - occupation-level AI exposure measures, including the AI substitution measure used for the treatment variable; contains fields AIsub, GBB, AIOE, CAIOE, GENOE1yr, GENOE5yr, GENOE10yr, read by 02_build_pooled_panel_for_replication.R lines 16/64; (2) emp_occ1.csv - occupation-level employment weights by country/year, read by 02_build_pooled_panel_for_replication.R lines 17/65; (3) wageworker2020_24.dta - income group classification (variable income_full_fy26) used to split countries into HIC/Non-HIC, read by 03_make_figure_4_7.R line 54, likely built from CLASS_2025_10_07.xlsx (see World Bank Country Income Classification entry). File location: "Chapter 4\1 - Data\Figure_4_7\Processed".
    Access policy
    Data is included in the reproducibility package
    Citation
    Authors' estimates. 2026. "Occupation and Income Classification Data" [dataset]. Based on multiple occupational AI exposure indices (Gmyrek/Berg/Bescond and Felten et al. measures) and World Bank income classifications.
    Data statement

    Some data is restricted and has not been included in the reproducibility package. For more details, refer to the README file.

    Description

    Output
    Type
    Flagships & Reports
    Title
    World Development Report 2026: The Promise of Artificial Intelligence
    Authors
    Author Affiliation Email
    Gaurav Nayyar World Bank gnayyar@worldbank.org
    Anukriti Siddharth World Bank sanukriti@worldbank.org
    Sharmista Appaya World Bank sappaya@worldbank.org
    Elwyn Davies World Bank edavies@worldbank.org
    Lelys Dinarte-Diaz World Bank ldinartediaz@worldbank.org
    Sam Fraiberger World Bank sfraiberger@worldbank.org
    Yi Jie Gwee World Bank ygwee@worldbank.org
    Yan Liu World Bank yanliu@worldbank.org
    Jeremy Ng World Bank njeremy@worldbank.org
    Tiago C. Peixoto World Bank tpeixoto@worldbank.org
    Manuel Ramos-Maqueda World Bank mramosmaqueda@worldbank.org
    Anja Sautmann World Bank asautmann@worldbank.org
    Katherine Stapleton World Bank kstapleton@worldbank.org
    Shu Yu World Bank syu2@worldbank.org
    Date of production

    2026-08-10

    Scope and coverage

    Geographic locations
    Location Code
    World WLD

    Disclaimer

    Disclaimer

    The materials in the reproducibility packages are distributed as they were prepared by the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this event do not necessarily reflect the views of the World Bank, the Executive Directors of the World Bank, or the governments they represent. The World Bank does not guarantee the accuracy of the materials included in the reproducibility package.

    Access and rights

    License
    Name URI
    MIT License https://opensource.org/license/mit
    World Bank IGO Rider https://github.com/worldbank/metadata-editor/blob/main/WB-IGO-RIDER.md

    Contacts

    Contacts
    Name Affiliation Email
    Gaurav Nayyar World Bank gnayyar@worldbank.org
    Reproducibility WBG World Bank reproducibility@worldbank.org

    Information on metadata

    Producers
    Name Abbreviation Affiliation Role
    Reproducibility WBG DECDI World Bank - Development Impact Department Verification and preparation of metadata
    Date of Production

    2026-08-10
    Document version

    1

    Citation

    Citation
    loading, please wait...
    Export citation: RIS | BibTeX | Plain text
    Back to Catalog