{"type":"script","doc_desc":{"producers":[{"name":"Reproducibility WBG","abbr":"DECDI","affiliation":"World Bank - Development Impact Department","role":"Verification and preparation of metadata"}],"prod_date":"2026-09-18","version":"1"},"project_desc":{"authoring_entity":[{"name":"Daniel Rogger","affiliation":"World Bank","email":"drogger@worldbank.org"},{"name":"Flavia Sacco Capurro","affiliation":"World Bank","email":"fsaccocapurro@worldbank.org"},{"name":"Manuel Ramos Maqueda","affiliation":"World Bank","email":"mramosmaqueda@worldbank.org"},{"name":"Timothy Lundy","affiliation":"World Bank","email":"tlundy@worldbank.org"},{"name":"Josefina Silva Fuentealba","affiliation":"World Bank","email":"jsilvafuentealba@worldbank.org"},{"name":"Shyam Jayanti Patel","affiliation":"World Bank","email":"spatel16@worldbank.org"}],"title_statement":{"title":"Reproducibility package for How is Government Using AI?","idno":"RR_WLD_2026_746"},"data_statement":"The reproducibility package relies on two types of data: open and limited-access (available to World Bank Staff).\n","software":[{"name":"R","version":"4.6.1"}],"scripts":[{"title":"Reproducibility package for How is Government Using AI?","date":"2026-09","notes":"Computational reproducibility verified by Development Impact (DECDI) Analytics team, World Bank.","instructions":"See README in reproducibility package.","file_name":"RR_WLD_2026_746","zip_package":"RR_WLD_2026_746.zip","dependencies":"R dependencies are listed in the file renv.lock."}],"repository_uri":[{"name":"Reproducible Research Repository (World Bank)","uri":"https:\/\/reproducibility.worldbank.org"}],"production_date":"2026-09-18","abstract":"Artificial intelligence (AI) has great potential to enhance public services. Yet there is little evidence on how governments are adopting AI, especially in non-high-income economies. This paper provides new cross-economy evidence from the AI and Data for Better Governance Survey, which covered 60 economies across all income levels and collected information from central digital agencies, sectoral ministries, and management information system leads.\nWe document three main findings. First, AI adoption by governments is broad but shallow and uneven. Most governments report some use of AI, yet adoption is generally ad hoc and rarely institutionalized. Second, constraints vary with income level.\nWhile a lack of data and insufficient budgets are common barriers everywhere, governments in low-income economies more often report a lack of policies, while those in high-income economies emphasize privacy and ethical concerns. Third, governance frameworks are lagging, with many governments providing access to generative AI tools before establishing formal guidelines.\nWe interpret these patterns through an ecosystem framework in which AI\u2019s value depends on complementary investments in data infrastructure, analytical capability, organizational capacity, and governance. We sort governments into four archetypes defined by how their AI adoption reflects their institutional readiness, including a few \u201ccowboys\u201d whose AI use is running ahead of their institutional scaffolding. The results suggest that AI policy for the public sector should shift from a narrow focus on adoption to a broader agenda of institutional readiness, responsible use, and sector-specific problem solving.","geographic_units":[{"name":"World","code":"WLD"}],"keywords":[{"name":"Artificial Intelligence"},{"name":"Public Sector"},{"name":"Digital Government"},{"name":"Governance"}],"output":[{"type":"Working Paper","description":"Policy Research Working Paper (PRWP)","title":"How is Government Using AI?"}],"language":[{"name":"English","code":"EN"}],"technology_requirements":"Runtime ~2 minutes.","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.","license":[{"name":"MIT License","uri":"https:\/\/opensource.org\/license\/mit"},{"name":"World Bank IGO Rider","uri":"https:\/\/github.com\/worldbank\/metadata-editor\/blob\/main\/WB-IGO-RIDER.md"}],"contacts":[{"name":"Daniel Rogger","affiliation":"World Bank","email":"drogger@worldbank.org"},{"name":"Reproducibility WBG","affiliation":"World Bank","email":"reproducibility@worldbank.org"}],"technology_environment":"Paper exhibits were reproduced on a computer with the following specifications:\n\u2022 OS: Windows 11 Enterprise\n\u2022 Processor: INTEL(R) XEON(R) Gold 5218 2.30 GHz (2 processors)\n\u2022 Memory available: 16.0 GB\n\u2022 Software version: R 4.6.1","reproduction_instructions":"To reproduce the findings in this paper, a replicator must:\n1. Open the R project (\"Reproducibility Package (August 28) 2.Rproj\") in RStudio.\n2. Run renv::restore() to recover the exact package versions used (see renv.lock), following the prompts.\n3. Open \"2 - Code\/AI Background Paper Reproducibility.Rmd\" and run it from beginning to end. No manual path edits are required \u2014 all file paths resolve automatically via here::here() relative to the project root.\n4. All figures generated by the code are written to \"3- Output\/\" (main-text Figures 1\u201319) and \"3- Output\/figures\/appendix\/\" (appendix Figures A.1\u2013A.6).","datasets":[{"name":"AI & Data for Better Governance Survey","note":"Filenames: Agency Questionnaire_WIDE.csv; Education Manager Survey.csv, Health Manager Survey.csv, Civil Service Manager Survey.csv, Taxation Manager Survey.csv, Public Finance Manager Survey.csv, Public Procurement Manager Survey.csv; Systems Questionnaire Education_WIDE.csv, Systems Questionnaire Health_WIDE.csv, Systems Questionnaire Civil Service_WIDE.csv, Systems Questionnaire Tax_WIDE.csv, Systems Questionnaire Public Finance_WIDE.csv, Systems Questionnaire Procurement_WIDE.csv. The survey has three components: (1) a whole-of-government (agency) questionnaire completed by each country's central digital authority or equivalent institution, covering AI adoption and the broader enabling environment; (2) manager questionnaires examining AI adoption, governance, and enabling conditions within the relevant ministry or agency; and (3) systems questionnaires capturing AI applications that use data from management information systems. Components 2 and 3 cover 6 core government sectors: tax, public finance, procurement, education, health, and civil service. For World Bank Staff, users can follow the link below and place the datasets in the correct folder. The placement of the needed datasets can be found in the data_hash_report.csv included in the reproducibility package. ","access_type":"Data is limited access and not included in the reproducibility package.","license":"Public Use Files","citation":"World Bank. 2026. \"AI & Data for Better Governance Survey\" [dataset]. World Bank Microdata Library, catalog 20124. https:\/\/microdatalib.worldbank.org\/index.php\/catalog\/20124","uri":"https:\/\/microdatalib.worldbank.org\/index.php\/catalog\/20124","license_uri":"https:\/\/microdatalib.worldbank.org\/index.php\/data-access\/#public"},{"name":"World Bank Country and Lending Groups","note":"Filename: OGHIST_2026_07_01_new.xlsx. World Bank GNI per capita Operational Guidelines & Analytical Classification used to assign economies to FY2026 income groups.","access_type":"Data is publicly available and included in the reproducibility package.","license":"Creative Commons Attribution 4.0 International License (CC BY 4.0)","citation":"World Bank. 2026. \"World Bank Country and Lending Groups\" [dataset]. Accessed 2026-08. https:\/\/datahelpdesk.worldbank.org\/knowledgebase\/articles\/906519-world-bank-country-and-lending-groups","uri":"https:\/\/datahelpdesk.worldbank.org\/knowledgebase\/articles\/906519-world-bank-country-and-lending-groups","license_uri":"https:\/\/www.worldbank.org\/ext\/en\/legal\/terms-conditions\/datasets"}]},"datacite":{"creators":[{"givenName":"Daniel","familyName":"Rogger","nameType":"Personal","affiliation":[{"name":"World Bank","affiliationIdentifier":"https:\/\/ror.org\/00ae7jd04","affiliationIdentifierScheme":"ROR","schemeUri":"https:\/\/ror.org"}]},{"givenName":"Flavia Sacco","familyName":"Capurro","nameType":"Personal","affiliation":[{"name":"World Bank","affiliationIdentifier":"https:\/\/ror.org\/00ae7jd04","affiliationIdentifierScheme":"ROR","schemeUri":"https:\/\/ror.org"}]},{"givenName":"Manuel Ramos","familyName":"Maqueda","nameType":"Personal","affiliation":[{"name":"World Bank","affiliationIdentifier":"https:\/\/ror.org\/00ae7jd04","affiliationIdentifierScheme":"ROR","schemeUri":"https:\/\/ror.org"}]},{"givenName":"Timothy","familyName":"Lundy","nameType":"Personal","affiliation":[{"name":"World Bank","affiliationIdentifier":"https:\/\/ror.org\/00ae7jd04","affiliationIdentifierScheme":"ROR","schemeUri":"https:\/\/ror.org"}]},{"givenName":"Josefina Silva","familyName":"Fuentealba","nameType":"Personal","affiliation":[{"name":"World Bank","affiliationIdentifier":"https:\/\/ror.org\/00ae7jd04","affiliationIdentifierScheme":"ROR","schemeUri":"https:\/\/ror.org"}]},{"givenName":"Shyam Jayanti","familyName":"Patel","nameType":"Personal","affiliation":[{"name":"World Bank","affiliationIdentifier":"https:\/\/ror.org\/00ae7jd04","affiliationIdentifierScheme":"ROR","schemeUri":"https:\/\/ror.org"}]}],"titles":[{"lang":"en","title":"Reproducibility package for How is Government Using AI?"},{"title":"RR_WLD_2026_746","titleType":"Other"}],"publisher":"World Bank","publicationYear":"2026","types":{"resourceType":"Reproducibility package","resourceTypeGeneral":"Other"},"url":"https:\/\/reproducibility.worldbank.org\/index.php\/catalog\/study\/RR_WLD_2026_746","language":"en"},"tags":[{"tag":"DOI"},{"tag":"Open Code"},{"tag":"Open Data"}],"schematype":"script"}