EU · France — an assembled comparison from each country's own governance data.
EU and France illustrate different points on the AI governance spectrum. EU's approach — Comprehensive, risk-based hard law binding all member states; France's approach — EU AI Act member state; AI sovereignty pursued mainly through infrastructure investment and certification rather than new national legislation.
On trajectory: EU is holding a steady regulatory posture; France is holding a steady regulatory posture. Most recent tracked marker for each: EU — "The AI Act's general-purpose AI (GPAI) supervision and enforcement powers take effect: the Commission's AI Office can now request…" (2026-08-02); France — "The EU AI Act's general-purpose AI (GPAI) supervision and enforcement powers take effect on 2 August 2026 — the AI Office can now request…" (2026-08-02).
Risk-tiered hard law, binding across all member states
EU AI Act member state; sovereignty pursued mainly through infrastructure investment and certification rather than new national legislation
EU (stable): The AI Act's core structure is unchanged, but June 2026's Digital Omnibus postponed several major compliance deadlines in response to competitiveness pressure — a real softening of pace even as the Commission simultaneously pushes forward on compute sovereignty via the newly enacted Cloud and AI Development Act. Its clearest legislative marker on file is "AI Act (Regulation (EU) 2024/1689), as amended by the Digital Omnibus" (in force). France (stable): France's substantive AI obligations remain those of the EU AI Act, whose GPAI supervision and enforcement powers activated on 2 August 2026; as an EU-based frontier lab, Mistral now falls directly under EU AI Office oversight. Nationally, France appears to have moved slowly on the mechanics of implementation — its September 2025 scheme designating competent authorities (with the DGCCRF as single point of contact) reportedly still awaited parliamentary adoption as of mid-2026. Most of the visible momentum has instead been industrial: a roughly €109B infrastructure package, Mistral's compute build-out and the MGX/Bpifrance/Nvidia 'Campus AI' venture, and efforts toward European silicon aimed at reducing NVIDIA dependence. Its clearest legislative marker on file is "National AI infrastructure package" (in force).
Compute sovereignty via CADA and 'AI Gigafactories'; simplification to aid competitiveness
Significant state investment in compute sovereignty via Mistral Compute, Jean Zay, and a further ~€4B French/Swedish data-centre build-out
EU: Very high — binding supranational law in force across all 27 member states, though national enforcement readiness reportedly varies (notable named developers: Mistral AI, Aleph Alpha). France: High — large state-backed infrastructure spend and a defense-AI contract with Mistral (notable named developers: Mistral AI, Hugging Face).
Participates in the G7 Hiroshima Process; the AI Act is frequently cited internationally as a reference point in AI regulatory debates
EU AI Act participation; ASML as Mistral's lead strategic investor
No shared "security"-tagged debate axis is recorded for both — on the evidence here, each is legislating its international posture on its own track rather than a shared one.
Balancing AI Act enforcement with industrial competitiveness and compute sovereignty
Apparent priority on sovereign compute and frontier-lab competitiveness, with limited new domestic AI-specific regulation beyond the EU AI Act
Where each government is actually spending its attention right now — not a statement of law, but of priority.
4 tracked measures (1 in force, 1 passed, 2 repealed)
4 tracked measures (3 in force, 1 proposed)
EU tracks 3 government AI institutions; France tracks 5 government AI institutions. Raw counts aren't a maturity score by themselves — a single binding statute can matter more than five proposals — but they mark how much formal apparatus each country has stood up so far.
Dedicated AI safety institute, Semiconductor and chip sovereignty
Frontier model developers, Advanced semiconductor fabrication
The gaps recorded for each country don't overlap — each is missing something different, which is a weaker signal than a shared gap but still worth reading as "not present as of the most recent research," not "definitively absent forever."
Shared
EU
France
EU and France each describe their own state's role differently: EU — "Very high — binding supranational law in force across all 27 member states, though national…"; France — "High — large state-backed infrastructure spend and a defense-AI contract with Mistral". Layered onto trajectory (EU stable, France stable), the open question this comparison actually poses is how much of each country's AI trajectory is being steered by the state versus left to market and research actors — and, for whichever side is tightening, whether that steering is increasing or easing right now.