Due to be unveiled Tuesday in Strasbourg by EU tech chief Henna Virkkunen, the plan is expected to address how to secure European access to cutting-edge — or “frontier” — AI models and explain how the bloc can impose its will on safety testing of advanced models.

Yet hanging over the Commission is an inescapably systemic problem that no action plan can overcome: The EU is dramatically behind the U.S. and China in developing frontier models and accessing the infrastructure that underpins them. Add to that the risk that Europe’s eagerness to have a voice on the safety of today’s leading AI models will only leave Europe more vulnerable in the long run.

The Commission “must avoid falling into a trap where pushing for access to the top-tier U.S. models could just deepen our dependency on American cloud infrastructure,” said Aura Salla, a Finnish lawmaker with the center-right European People’s Party.

European policymakers have fretted for years about the dominance of U.S. tech and whether — or when — U.S. President Donald Trump might hit the so-called kill switch to cut off access to essential digital platforms and tools.

Mythos, which Anthropic claimed has unprecedentedly powerful cyber capabilities, quickly became an avatar for those concerns in the AI era.

First, there were no Europeans on the original April list of companies and institutions invited to test Mythos for cyber risks. Then last month, the U.S. government blocked Anthropic from exporting its latest AI models. That ban has since been partially lifted.