Does CADA fund AI factories and gigafactories?

Summary The proposed Cloud and AI Development Act (CADA) does not establish a direct funding line to construct AI factories or gigafactories. Instead, it creates a mandatory strategic framework requiring Member States to plan and invest in these high-intensity compute infrastructures. Specifically, Article 7(2)(e) obliges national cloud and AI strategies to include "measures to invest in high-intensity computing infrastructure, including AI factories, AI gigafactories and quantum computers." While CADA sets the regulatory requirement for these investments, the actual financing is expected to come from national budgets, state aid, and complementary EU instruments like the European Competitiveness Fund. This national planning requirement works in tandem with Article 9, which provides Union-level compute support specifically for designated "frontier AI priority projects," creating a two-tiered approach to expanding Europe's compute capacity.

Detail

The Cloud and AI Development Act (CADA), as proposed in COM(2026) 502 final, represents a significant shift in how the EU approaches digital infrastructure. Rather than acting as a direct construction fund for specific data centres or AI factories, the proposal functions as a regulatory and strategic coordination instrument. Its primary mechanism for driving the deployment of AI factories and gigafactories is the imposition of binding planning obligations on Member States, coupled with a targeted Union-level matching mechanism for specific high-priority projects.

The Mandatory Planning Obligation: Article 7

The core of CADA's approach to AI factories lies in Article 7, titled "National cloud and AI strategies." Under Article 7(1), Member States are required to establish these national strategies within one year of the Regulation's entry into force. These strategies are not merely advisory; they must adhere to a strict minimum content framework defined in Article 7(2).

Crucially, Article 7(2)(e) explicitly mandates that national strategies must include:

"measures to invest in high-intensity computing infrastructure, including AI factories, AI gigafactories and quantum computers as strategic national and cross-border assets supporting research, development and industrial AI deployment across strategic sectors;"

This provision is transformative for the EU's compute landscape. By embedding "AI factories" and "AI gigafactories" directly into the statutory text of the national strategy requirements, CADA elevates these facilities from optional industrial projects to essential components of national digital sovereignty. The text further clarifies that these assets are to be viewed as "strategic national and cross-border assets," implying that their deployment is a matter of Union-wide interest, not just local economic development.

The requirement is specific: Member States cannot simply acknowledge the need for compute; they must outline concrete "measures to invest." This forces national governments to allocate resources, design public-private partnership models, or create regulatory sandboxes to facilitate the construction of these high-intensity facilities. The inclusion of "quantum computers" alongside AI factories and gigafactories in Article 7(2)(e) signals a holistic approach to next-generation computing, ensuring that national strategies address the full spectrum of future computational needs.

Complementarity with Union Compute Support: Article 9

While Article 7 drives the infrastructure planning at the national level, Article 9 addresses the allocation of compute resources at the Union level. This creates a complementary ecosystem where national governments build the capacity, and the Union ensures it is utilized for strategic priorities.

Article 9, titled "Computing support for AI projects," focuses on "frontier AI priority projects." Under Article 9(1), the Union and Member States must ensure that sufficient AI computing resources are allocated to support these designated projects, "within the limits of available capacity." This capacity could theoretically come from the very AI factories and gigafactories that Member States are mandated to build under Article 7.

The mechanism for Union support is detailed in Article 9(2), which states:

"The Union shall at least match the AI computing resources contributed by Member States to frontier AI priority projects to the extent that sufficient AI computing capacity is available within the Union's share of European high performance computing access time."

This "matching" provision is a powerful lever. It incentivizes Member States to not only build infrastructure but to actively contribute compute time from that infrastructure to Union-level strategic projects. The Union's contribution is drawn from its share of European High Performance Computing (EuroHPC) capacity. This creates a feedback loop: national investment in AI factories (Article 7) increases the pool of available capacity, which in turn allows the Union to provide more matching resources for frontier AI projects (Article 9).

It is important to note that Article 9 support is not universal. It is strictly tied to projects that meet the criteria for "frontier AI priority projects" as defined in Article 8. These criteria include being a "pioneering project" focused on scaling up frontier AI technologies and involving broad participation from entities across the Union. Therefore, while Article 7 ensures the existence of AI factories, Article 9 ensures that a portion of their capacity is directed toward the most strategically critical AI development efforts.

The Funding Reality: Regulatory Mandate vs. Financial Instrument

A common point of confusion is the source of funding. The text of CADA makes a clear distinction between the obligation to plan and invest and the source of the funds.

CADA itself does not contain a dedicated budget line for the construction of AI factories. The proposal is a Regulation, which primarily sets rules and frameworks. The explanatory memorandum and the Legislative Financial Statement clarify that the Cloud and AI Leadership Initiatives (Title II) may be supported by funding from existing Union programmes, such as Horizon Europe and the Digital Europe Programme. Furthermore, the proposal explicitly references the future European Competitiveness Fund (ECF) as a key financial driver.

The "measures to invest" required by Article 7(2)(e) are expected to be financed through a combination of:

National Budgets and State Aid: Member States will use their own fiscal resources, subject to EU state aid rules, to fund the construction of these strategic assets.
The European Competitiveness Fund (ECF): The proposal anticipates that the ECF, particularly under its "Digital Leadership" window, will provide significant deployment support.
InvestEU and Cohesion Policy: These instruments can mobilize private investment and support regional disparities in compute capacity.
Private Sector Investment: The regulatory certainty provided by CADA is designed to de-risk private investment in AI factories and gigafactories.

Thus, CADA acts as the "architect" of the investment landscape, forcing the blueprint (national strategies) and aligning the Union's matching resources (Article 9), while the "builders" (national treasuries, the ECF, and private capital) provide the bricks and mortar.

What this means for you

For technology leaders, investors, and policy advisors, the distinction between CADA's regulatory mandate and its financial mechanisms is critical for strategic planning.

For CTOs and Infrastructure Architects: You should anticipate a rapid formalization of high-intensity compute availability across the EU. Because Article 7(2)(e) requires Member States to include AI factories and gigafactories in their national strategies within one year of entry into force, you can expect a wave of national announcements detailing specific investment plans and locations for these facilities. When planning your infrastructure strategy, you should monitor the national strategies of the Member States where you operate. These documents will reveal where sovereign, high-performance compute capacity will be concentrated. This allows you to plan for low-latency access to these resources and potentially leverage the "EuroCloud Federation" or joint procurement mechanisms established under CADA to access them. The matching mechanism in Article 9 suggests that if your organization is involved in a "frontier AI priority project," you may have access to Union-matched compute resources derived from these national investments.

For Investors and Developers: The mandatory nature of Article 7(2)(e) reduces the regulatory risk associated with investing in AI factories. Member States are legally bound to create the conditions for these investments, creating a more predictable market environment. However, investors must understand that CADA does not provide direct grants for construction. The funding will likely come from the ECF or national co-financing. Therefore, investment cases should be built around the availability of these complementary funding streams rather than expecting CADA to be the primary funder. The focus on "quantum computers" alongside AI factories in the same clause also suggests that diversified compute portfolios (AI + Quantum) may be favored in national strategies.

For SMEs and Startups: While SMEs are unlikely to build AI factories, Article 7(2)(e) ensures that these facilities are designed to support "industrial AI deployment across strategic sectors." This implies that the capacity generated will be accessible to a broader range of industries, not just large hyperscalers. SMEs should look to the "Experience and Acceleration Centres for AI" (Article 5) as potential gateways to accessing these new resources. Furthermore, if an SME is part of a consortium developing a "frontier AI priority project" (Article 8), it may benefit from the Union's matching compute resources under Article 9, effectively lowering the cost of training and development.

Common misconceptions

Misconception 1: CADA is a direct funding program for AI factories. This is incorrect. CADA is a regulatory framework. It mandates that Member States plan to invest in AI factories and gigafactories via Article 7(2)(e), but it does not allocate a specific budget to build them. The actual funding comes from national budgets, the European Competitiveness Fund, and private investment. CADA provides the strategic alignment that makes these investments eligible for other funding streams.

Misconception 2: All AI projects get Union compute support. The compute matching mechanism in Article 9 is highly targeted. It applies only to "frontier AI priority projects" that meet the strict criteria of Article 8 (e.g., pioneering nature, broad Union participation). It does not apply to general commercial AI projects or the routine operation of AI factories. The Union's matching is a strategic lever for specific high-impact initiatives, not a general subsidy for all compute usage.

Misconception 3: AI factories are solely for large tech companies. While AI factories are large-scale, Article 7(2)(e) explicitly links them to "industrial AI deployment across strategic sectors." This includes sectors like healthcare, manufacturing, and energy, where SMEs are often the primary innovators. The goal of the mandate is to ensure that high-intensity compute is available to support the broader European industrial base, not just to serve the needs of a few large technology providers.

This is general information about a draft EU regulation, not legal advice.