United States: critical partner, not for new European fab capital

The U.S. is indispensable to the EU in five dimensions:

• x86 architecture (Intel, AMD)
• AI models (OpenAI, Anthropic, Google, Meta)
• Hyperscale cloud (AWS, Azure, GCP)
• Cybersecurity technologies (Palo Alto, CrowdStrike, Microsoft Defender)
• Research collaboration (MIT, Stanford, DARPA, joint NSF projects)

But the CHIPS Act has redirected American investment capital toward domestic projects (TSMC Arizona — now a $165 bn commitment , Samsung Taylor TX, Intel Ohio and Oregon). The EU has nothing comparable to offer — not at the level of $52 bn in direct subsidies plus a 25 % advanced-manufacturing investment tax credit. Hoping for transatlantic investment flows in the coming three years is not realism.

Realistic approach: maintain regulatory and defence cooperation; accept that American capital will not build new fabs in Europe in the foreseeable horizon; concentrate energy on contractual isolation of extraterritorial risks through the operational checklist.

The cancellation of Intel Magdeburg in July 2025 was the watershed event. Before that cancellation, the partnership map could include U.S. capital as a realistic source of European industrial investment. After it, that assumption is wishful thinking. Strategic clarity requires accepting this and reallocating attention toward Korea, Japan, and selectively other partners.

The U.S. relationship is critical and strategic — but it is no longer about new European industrial capacity. It is about technology access, regulatory coordination, and shared defence.