On July 15, 2026, the European Commission’s AI Office published a report summarizing findings from more than 100 experts drawn from European AI developers, industry, investors, academia, think tanks, member states, and the Commission itself on how the EU can strengthen its competitiveness, technological sovereignty, and security in frontier AI. It is a notable document because it comes from the same office that enforces the AI Act, and its diagnosis is blunt.
The experts find that frontier AI capabilities have advanced at an unprecedented pace - moving in roughly three years from struggling with basic tasks to approaching the limits of what current tests can measure - while Europe’s position remains modest relative to its economic weight. Frontier development is concentrated outside the EU, and the report is explicit that non-frontier models are an imperfect substitute. The experts caution that “the next one to two years could be decisive for Europe to move to a position of strength.”
The recommendations are organized around removing barriers to competitiveness, with access to advanced computing infrastructure and the energy to power it named as immediate priorities for the next two years; addressing growth-stage capital constraints and clarifying copyright and data-protection rules for training data; strengthening the EU’s sovereign ability to access, select, control, and benefit from frontier models through partnerships with trusted allies; and setting ambitious 2030 targets backed by real institutional capacity.
For business leaders, the report signals where EU AI policy energy is shifting: after two years dominated by the AI Act’s obligations, the bloc’s own AI Office is now publicly framing compute, capital, and legal clarity for training data as urgent competitiveness problems - a framing likely to shape both the pending AI Act simplification work and EU infrastructure spending.