Richer than their data: an income-adjusted panel analysis of the statistical foundations for health AI in Arab League states, 2016–2024

Background. Artificial intelligence (AI) for health depends on national data systems that are timely, open and well-governed. Arab League states, especially the Gulf Cooperation Council (GCC), are investing heavily in health AI, but no study has asked whether their statistical foundations match what their income predicts. Methods. We linked the World Bank Statistical Performance Indicators (SPI) for 2016–2024 to GDP per capita, population and governance indicators. The primary outcome was the SPI overall score (0–100); secondary outcomes were the five SPI pillars and five health-relevant items. Pooled OLS with year fixed effects and country-clustered errors estimated the income-adjusted difference between Arab League members and other countries, with wild cluster bootstrap inference and nine sensitivity analyses (STROBE/RECORD). Results. We analysed 1,586 country-years from 185 countries (21 Arab League members). Arab League states scored 15.1 points below what income and population predict (95% CI −21.0 to −9.3; bootstrap p<0.001). The largest pillar gap was in data infrastructure (−19.0). The gap was largest in the GCC (−22.3; infrastructure −33.0; data services −27.7). It narrowed by 1.2 points per year, from −20.2 in 2016 to −11.3 in 2024. Health-specific items showed no robust gap. Adjusting for voice and accountability reduced the overall gap by 58%. Conclusions. The Arab region's constraint on health AI lies less in missing health indicators than in under-built, less open statistical systems around them, particularly in the high-income GCC. Health-AI investment should include data infrastructure and services.