Does Sovereign AI Foster Innovation? Evidence from Global Data Governance Regimes

Abstract: Artificial intelligence (AI) is increasingly recognized as a strategic general-purpose technology shaping economic competitiveness, national security, and technological autonomy. The emerging notion of sovereign AI captures governments’ efforts to secure control over critical AI enablers—data resources, compute infrastructure, cloud platforms, foundational models, and skilled human capital. Simultaneously, states adopt diverse data governance regimes ranging from comprehensive data protection laws and localization requirements to open cross-border data flows. This paper examines how sovereign AI and data governance relate to innovation performance across countries. Using a cross-country quantitative analysis (2011–2023) for 56 countries, we construct composite indicators of sovereign AI capacity and data governance restrictiveness, linking them to measures of national innovation output. Panel regression models controlling economic structure, human capital, and digital infrastructure reveal that higher sovereign AI capacity is positively associated with innovation performance, especially when paired with calibrated data governance frameworks that protect personal data while enabling industrial data flows. Conversely, restrictive localization policies without strong domestic AI capabilities are not consistently linked to better innovation outcomes. Findings suggest that effective sovereign AI strategies require balanced investment in domestic AI infrastructure and calibrated governance regimes that safeguard rights without fragmenting data ecosystems essential for innovation.