Uneven Use, Similar Trust Generative AI in Denmark and Romania
Generative AI (GenAI) has spread unevenly across Europe, and it is unclear whether higher use brings higher trust. This paper compares Denmark and Romania, the EU's highest and lowest GenAI users, and asks whether the difference in public and institutional trust between them corresponds to the difference in trust in AI. Drawing on trust, technology-acceptance and digital-divide theory, it is a descriptive, contrasting-case analysis of harmonised data: GenAI usage comes from Eurostat, trust in AI more generally from the KPMG/Melbourne global study and public and institutional trust from the Standard Eurobarometer. In 2025, 48.4% of Danes aged 16-74 had used GenAI in the past three months, against 17.8% of Romanians. Use falls with age and rises with education in both countries, and the relative gaps are far larger in Romania; the patterns remain visible within the published age-by-education categories. Trust in AI does not follow usage: Romanians trust AI as much as Danes or slightly more, although Denmark leads by 28 to 45 percentage points on trust in government, parliament and the legal system. The Denmark-Romania comparison thus does not support a simple country-level correspondence between general institutional trust and trust in AI; in KPMG's pooled model, trust in AI is tied most closely to confidence in AI governance. The observed usage difference is more consistent with disparities in general digital skills and opportunities for work- and education-related use than with differences in trust. However, the available data cannot identify the independent effects of these factors. The results rest on two cases and non-identical constructs, and call for comparable microdata and disclosure experiments.