FEATURED

There’s a number buried in this story that nobody in U.S. equity research is taking seriously.

The scale of investment required under Europe’s new tech sovereignty program is not small: an estimated €120 billion for semiconductors, €200 billion for data centers by 2036, €100 billion for cloud and AI, and €2 billion for open-source software over seven years. Add it up and you’re looking at roughly €420 billion in directed European spending aimed at building domestic alternatives to infrastructure that U.S. companies currently own. That’s not a European subsidy story. That is a direct competitive threat to the revenue base of every major U.S. cloud and chip company operating in Europe.

On June 3, 2026, the European Commission put forward the European Technological Sovereignty Package, a set of measures to strengthen the EU’s capacity in semiconductors, artificial intelligence, cloud computing, and open source. On the surface it reads like industrial policy. What it actually is: a procurement exclusion mechanism wrapped in sovereignty language.

What’s Actually in It

The package includes two major legislative proposals: Chips Act 2.0 and the Cloud and AI Development Act (CADA). Both outline sweeping objectives for the EU’s semiconductor industry and local cloud and AI providers. But the Cloud and AI Development Act is the sharper instrument.

The CADA introduces a formal sovereignty assessment into public procurement for the first time. Member states would be required to evaluate the jurisdictional risks of their digital contracts before signing them. At the most sensitive tier, covering critical government systems, U.S. cloud hyperscalers would be barred outright from competing. The Act creates four sovereignty tiers. Only European providers, or non-European ones that can demonstrate verified data localization and legal ring-fencing, would qualify for the top two.

Read that slowly. It’s not a fine. It’s not a negotiation. It’s a structural bar baked into procurement law.

Chips Act 2.0 is the other half. It grants the Commission emergency powers to override semiconductor firms’ commercial contracts during a shortage and compel them to prioritize EU crisis-critical orders. Companies complying would be shielded from liability for breaching existing contracts. The Commission is also targeting €120 billion in semiconductor investment by 2035 and is planning the EU’s first advanced chip manufacturing foundry, with pilot production envisaged between 2030 and 2033.

Here is what makes this different from every prior EU digital regulation: it combines demand-side pressure with supply-side subsidy in one package. The Commission seeks to reduce supply-side dependencies by strengthening domestic capabilities and stimulating demand in downstream sectors, spanning the full tech stack from chips and infrastructure to software, cloud, and AI.

The Risk Equity Models Aren’t Catching

The Commission’s own communication states the EU remains structurally reliant on non-EU providers for more than 80% of digital products, services, infrastructure, and intellectual property. Europe produces only around 10% of global semiconductors. More than 70% of the EU cloud market is held by three U.S. hyperscalers: AWS, Azure, and Google Cloud. Those three are the target. The entire addressable market being contested sits inside their current revenue base.

Here’s the structural problem that nobody in sell-side research wants to say plainly: it would be difficult for U.S. companies to reach the highest levels of sovereignty because of the U.S. Cloud Act. That law allows U.S. law enforcement to request user data from American companies regardless of where the data is stored, including on servers physically located in Europe. The Commission is now trying to turn that legal exposure into a procurement disqualifier. As one EU official put it, “We want to make sure that our most critical sensitive data is stored in Europe.”

No amount of data center investment in Frankfurt changes the underlying legal reality of the Cloud Act. That’s a structural disqualification, not a compliance checkbox U.S. hyperscalers can write their way around.

Slight tangent, but it matters: the Microsoft situation earlier this year accelerated this entire conversation inside Brussels. When the Trump administration sanctioned the International Criminal Court’s top prosecutor and Microsoft subsequently canceled his email account, it crystallized the “kill switch” concern that EU policymakers had been abstractly worried about for years. That single episode gave the sovereignty package political momentum it might not otherwise have had.

On the chip side, the AI chip market is estimated at around $100 billion in 2026 and is projected to nearly triple to roughly $293 billion by 2030. NVIDIA holds well over 80% of the current AI GPU market, and its data center revenue is projected to rise from approximately $115 billion in 2025 to around $483 billion by 2030. The EU’s revised Chips Act shifts focus from supply to demand, seeking to connect European chipmakers with domestic industrial users. If that works, even partially, it redirects significant procurement spending away from NVIDIA, AMD, and Intel toward European alternatives. The EU also plans a major expansion of data center capacity, aiming to roughly triple output within five to seven years.

The timeline for legislative passage is not fast. Both the Chips Act 2.0 and the CADA must still be negotiated and approved by the European Parliament and the Council of the European Union. The original Chips Act took roughly two years from proposal to adoption. CADA, which touches the politically sensitive question of digital sovereignty vis-a-vis key trading partners, may take longer. That is exactly why U.S. equity models are ignoring it. Markets tend to price regulatory risk at the moment of enforcement, not the moment of proposal. The gap between those two points is where the asymmetry sits.

The stocks most exposed: Amazon (AWS European revenue), Microsoft Azure, which leads the European public sector cloud market, and Alphabet Cloud. The stocks that potentially benefit: ASML, which already dominates EUV lithography, Infineon Technologies, and STMicroelectronics, all positioned as the domestic European chip supply chain the Commission is explicitly trying to build. None of that is fully reflected in current valuations. The question isn’t whether this policy matters. It’s whether you want to wait until enforcement to find out.