The EU AI Act Newsletter #100: The European Way

Key Takeaways

• 19 AI Factories now operating on European supercomputers
• 76 expressions of interest for AI Gigafactories across 16 states
• EU AI Office seeks $109 million yearly, 100 staff by 2030
• Anthropic’s Mythos model inaccessible to most European agencies
• Annex I merger consolidates compliance pathways without deregulation

Pulse Analysis

The European Union is cementing its AI agenda through a multi‑pronged approach that blends infrastructure, funding, and regulatory refinement. The AI Continent Action Plan’s rollout of 19 AI Factories and the burgeoning interest in AI Gigafactories illustrate a concerted effort to build a sovereign computing backbone. Coupled with roughly $1.09 billion earmarked for high‑impact sectors—such as AI‑driven cancer screening—the EU aims to nurture home‑grown innovation while maintaining a risk‑aware stance. This momentum is further reinforced by the AI Omnibus, which seeks to tighten the Act’s timeline ahead of the August 2026 high‑risk obligations deadline.

Academic and industry voices are now dissecting the EU’s General‑Purpose AI (GPAI) provisions, offering a nuanced map of legal certainty and ambiguity. The Cambridge Commentary, produced by the Leverhulme Centre and the Institute for Law & AI, provides a systematic analysis of each GPAI clause, helping stakeholders anticipate compliance pathways. Meanwhile, technologists like Kleitia Zeqo argue that the EU’s single‑market framework and risk‑based approach give Europe a structural edge over the fragmented U.S. regulatory landscape, positioning the AI Act as an enabler rather than a brake on innovation.

Despite these advances, practical challenges threaten to stall progress. Anthropic’s decision to limit its Mythos model to a handful of U.S. firms leaves European cyber agencies without critical testing capabilities, raising alarms about private governance of high‑risk AI. Simultaneously, a coalition of civil‑society groups is pressing the EU to allocate about $109 million annually and expand the AI Office’s staff to at least 100 by 2030, citing the DSA’s $55 million budget for 160 personnel as a benchmark. The recent Annex I merger, clarified by DIGITALEUROPE, aims to streamline compliance without diluting safety standards, underscoring the EU’s commitment to a balanced, competitive AI ecosystem.