Global country-level dataset and analysis code for AI readiness and ecosystem-service mismatch using the Oxford Insights Government AI Readiness Index 2025

This record contains the processed country-level dataset and R analysis code used for the study on global AI readiness and ecosystem-service mismatch. The dataset integrates World Bank indicators, the Yale Environmental Performance Index 2024, and the Oxford Insights Government AI Readiness Index 2025 to evaluate whether national AI capacity is aligned with ecosystem-service dependence, benefit-capture deficits, and environmental burden. The package includes a cleaned country-year panel for 2015 to 2025, a latest-available country cross-section, Oxford 2025 AI readiness variables, World Bank metadata, EPI 2024 selected indicators, country matching diagnostics, source maps, and a reproducible R analysis script. Derived indices include ecosystem-service potential, ecosystem-service dependence, ecosystem-service benefit capture, AI need, AI capacity, and AI–ecosystem-service mismatch. The analysis code implements country-level model fitting, mismatch diagnostics, sensitivity analyses, digital-divide diagnostics, residual AI-readiness analyses, inequality decomposition, permutation tests, measurement-coverage analysis, and figure-ready outputs. The Oxford 2025 report provides official ranks and six pillar scores. The variable oxford_gair_2025_pillar_mean is a package-derived arithmetic mean of the six published pillar scores for modelling convenience and should not be interpreted as an official Oxford aggregate score. Raw third-party downloads and report PDFs are not redistributed in this public-sharing package. Users should obtain third-party raw data from the original providers where required by source-provider terms.