Study: AI diffusion and investment stronger domestically

The research involved a large-scale field experiment with 3,300 firms across twelve European Union countries, conducted in the fourth quarter of 2025.

Firms were found to substantially underestimate competitors' current AI investment, both domestically and abroad, with an average gap exceeding twenty percentage points.

Providing accurate information about peer investment caused firms to update their expectations and significantly increase their own expected AI investment rate.

Crucially, the study identifies strong strategic complementarities within borders: a one-percentage-point increase in the expected share of domestic peers investing in AI raises a firm's own expected AI investment rate by approximately 0.57 percentage points.

These complementarities are entirely absent across borders, as an increase in the expected share of foreign peers investing in AI had no statistically significant effect on a firm's own investment plans.

This asymmetry holds across various robustness checks and subsamples, underscoring the dominant role of domestic competitive pressure.

The findings suggest that informational frictions, like firms underestimating peer adoption, contribute to underinvestment in AI.

The sharp asymmetry between domestic and foreign competitive pressure implies limited cross-border innovation spillovers, even within the Single Market.

This helps explain why regional disparities in AI adoption persist, as firms primarily respond to domestic peers.

From a policy perspective, financial incentives alone may not suffice to accelerate AI diffusion.

Complementary measures, such as information campaigns to improve firms' awareness of peer adoption, could be cost-effective tools to stimulate investment, especially where adoption lags.

The study underscores the continued importance of domestic business environments for innovation convergence.