EU AI Office outlines frontier AI priorities for competitiveness and security

The AI Office has published findings from over 100 experts on how the European Union can improve competitiveness, sovereignty and security in frontier AI. The report draws on the first meeting of the Expert Forum on Frontier AI, held in April, and related written statements as the EU works toward becoming an AI continent.

Experts said frontier AI has advanced at an unprecedented pace, moving within roughly three years from struggling with basic tasks to nearing the limits of current tests. They warned that Europe’s position remains modest compared with its economic weight, despite world-class research and a significant share of global AI talent, with frontier development still concentrated largely outside the EU.

Recommendations focused on removing structural barriers to attractive market conditions, with computing infrastructure and the energy to power it described as the most urgent two-year priorities. Experts also cited growth-stage capital, legal certainty around training data in copyright and data protection, and talent attraction and retention as areas needing action, while warning that the next one to two years could be decisive.

The AI Office will group recommendations by priority and continue expert exchanges through written input and targeted workshops. Experts also urged ambitious targets for 2030 and called for frontier AI to become a political priority backed by capacity and coordination across institutions.