In the last decade, countries have been struggling to secure their competitive edge in the burgeoning field of Artificial Intelligence (AI). In this paper, we discuss some key policy drivers that may help national governments keep pace with the global AI race. Additionally, given the rising cooperative and competitive dynamics emerging at the international level, we investigate the spillovers between countries. The findings highlight that regulation, talent, data availability and infrastructure are key enablers of AI deployment but also reveal the risks of misaligned policies and excessive intervention. For governments, the challenge lies in avoiding overinvestment while crafting the balanced policy frame. Spatial econometric modelling provides insights into scientific and technical knowledge flows (respectively, among AI researchers and practitioners). The positive spillovers of AI talent, as well as STEM and digital skills, through the network of scientific research underline the role of scientific cross-border cooperation. Conversely, the negative spillovers from technological capacity suggest that strengthening national AI systems may jeopardize collaborative research efforts. Similarly, data availability facilitates technical knowledge exchange, while the negative spillovers of infrastructure indicate that investment in it exacerbates competition. Findings have strong policy implications
Artificial intelligence and competitive edges: Do knowledge spillovers undermine strategic autonomy?
Marra, Alessandro
;Cartone, Alfredo;
2026-01-01
Abstract
In the last decade, countries have been struggling to secure their competitive edge in the burgeoning field of Artificial Intelligence (AI). In this paper, we discuss some key policy drivers that may help national governments keep pace with the global AI race. Additionally, given the rising cooperative and competitive dynamics emerging at the international level, we investigate the spillovers between countries. The findings highlight that regulation, talent, data availability and infrastructure are key enablers of AI deployment but also reveal the risks of misaligned policies and excessive intervention. For governments, the challenge lies in avoiding overinvestment while crafting the balanced policy frame. Spatial econometric modelling provides insights into scientific and technical knowledge flows (respectively, among AI researchers and practitioners). The positive spillovers of AI talent, as well as STEM and digital skills, through the network of scientific research underline the role of scientific cross-border cooperation. Conversely, the negative spillovers from technological capacity suggest that strengthening national AI systems may jeopardize collaborative research efforts. Similarly, data availability facilitates technical knowledge exchange, while the negative spillovers of infrastructure indicate that investment in it exacerbates competition. Findings have strong policy implicationsI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


