Summary "AI factories" are large-scale, high-performance computing facilities meant to give European businesses and researchers broad access to next-generation compute for AI. As proposed, the Cloud and AI Development Act (CADA) does not define AI factories in its definitions article and does not create a dedicated fund for them. It treats them as strategic infrastructure: Article 7(2)(e) requires Member States to include measures to invest in high-intensity computing infrastructure — explicitly naming AI factories — in their national cloud and AI strategies. Funding would come from existing or planned Union instruments (notably the Digital Europe Programme and Horizon Europe under Article 6(3), plus InvestEU and the future European Competitiveness Fund), together with national support under State aid rules and private investment.

Detail

CADA aims to expand the EU's cloud and AI compute capacity and reduce reliance on a small number of non-EU hyperscalers. The explanatory memorandum notes that "the ongoing deployment of AI factories and AI gigafactories aims to provide broad access to high-capacity, next-generation computational resources for European businesses and researchers requiring AI capabilities." AI factories are part of that capacity push.

What are AI factories?

CADA does not give a standalone technical definition of "AI factory" in Article 2 (its definitions article). The term appears in the explanatory memorandum and once in the enacting text, where it is grouped with other high-intensity computing infrastructure. The concept refers to facilities focused on high-performance, AI-optimised computing — the kind needed to train, fine-tune and run large AI models — as distinct from general-purpose data centres. The explanatory memorandum frames them within the European AI Continent Action Plan's "nexus" of five domains: computing infrastructures, data, skills, development and adoption of AI algorithms, and regulatory simplification.

Strategic status under Article 7

The clearest legal hook in CADA's enacting text is Article 7, on national cloud and AI strategies. Member States must establish these strategies within one year of entry into force (Article 7(1)). Article 7(2)(e) requires them to 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 elevates AI factories from ordinary commercial facilities to strategic national and cross-border assets that Member States are expected to plan for and invest in. The reference to "cross-border assets" signals that facilities may be jointly planned or shared between Member States.

How are AI factories funded?

CADA does not establish an "AI Factory Fund." It provides the strategic framework and channels financing through existing and upcoming instruments:

  1. Union programmes (Article 6(3)). The Cloud and AI Leadership Initiatives "may be supported by funding from Union programmes, including Horizon Europe and the Digital Europe Programme" (Article 6(3)). The Digital Europe Programme is a primary vehicle for deploying compute infrastructure; Horizon Europe (and its successor, FP10) supports the upstream research and innovation dimension.
  2. InvestEU. Recital 28 adds InvestEU to the list of instruments that may support the initiatives, helping improve the investment environment and mobilise public and private capital.
  3. The European Competitiveness Fund (ECF). The explanatory memorandum expects the ECF — still a proposal — to serve as the main deployment instrument under the post-2027 multiannual financial framework, especially under its Digital Leadership window.
  4. National support and State aid. Recital 29 allows Member States to support the initiatives through research, development and innovation measures "in line with the applicable State aid rules."
  5. Private investment. Recital 29 also encourages private-sector stakeholders to align their cloud and AI investment strategies with the Leadership Initiatives, creating a coordinated investment pathway.

A note on EuroHPC: the European High Performance Computing Joint Undertaking is closely linked to this ecosystem. Article 9(2) provides that, for recognised frontier AI priority projects, the Union shall at least match the AI computing resources contributed by Member States "to the extent that sufficient AI computing capacity is available within the Union's share of European high performance computing access time." Recital 26 notes that implementation of the Leadership Initiatives could be entrusted to joint undertakings such as the EuroHPC JU.

Where AI factories fit in the Leadership Initiatives

AI factories are best understood as one output of the broader Cloud and AI Leadership Initiatives (Articles 3 to 6), whose operational objectives are set out in Article 4. Several of those objectives bear directly on the kind of capability AI factories provide: developing AI-optimised servers and baseline software based on processors and accelerators designed and manufactured in the Union (Article 4(2)(b)); boosting data availability for AI (Article 4(2)(c)); supporting sectoral AI models across strategic industrial sectors and facilitating "access to the necessary computing resources and AI tools" required to develop them (Article 4(5)); and, under operational objective 8, supporting the procurement of data centre and cloud computing services for Union entities and public sector bodies (Article 4(8)(d)). Funding routed through these objectives is therefore the practical channel by which AI-factory capacity is built and made accessible.

Sustainability by design

CADA ties future compute infrastructure to sustainability. The Cloud and AI Leadership Initiatives' first operational objective (Article 4(1)) is to advance energy- and water-efficiency technologies for data centres, including innovative cooling, next-generation direct-current data centres, waste-heat utilisation and integration with energy grids. Article 4(2) targets secure, resilient open cloud computing stacks. In practice, public support for AI-factory-type infrastructure is likely to favour facilities that meet high sustainability standards, consistent with these objectives.

AI factories versus AI gigafactories

CADA names AI factories and AI gigafactories side by side in Article 7(2)(e) but treats them as different points on the same spectrum of high-intensity computing infrastructure. The working distinction, drawn from the explanatory memorandum and surrounding EU policy, is one of scale and purpose: AI factories are oriented to providing broad access to next-generation compute for European businesses and researchers across many use cases, while gigafactories are the largest facilities aimed at training the most advanced, frontier AI models. Both are designated as strategic national and cross-border assets and both rely on the same funding architecture — Union programmes, national support under State aid rules, and private investment — rather than on any dedicated CADA appropriation. Neither term is defined in Article 2, so the boundary between them is descriptive rather than legal.

What this means for you

For CTOs, architects and SMEs:

  • Access to European compute. CADA aims to make high-capacity, next-generation compute available within the EU, reducing reliance on non-EU hyperscalers for training and fine-tuning workloads.
  • Watch your national strategy. Concrete funding calls and access programmes will surface through your Member State's Article 7 strategy, which must address where it invests in high-intensity computing infrastructure.
  • SME access is an explicit aim. The proposal seeks "broad access" for European businesses and researchers; Article 33 directs Member States to monitor and improve SME participation in cloud and AI procurement of innovation (with an objective of at least 25% of such procurement going to innovative SMEs).
  • Build for efficiency. Given the emphasis in Article 4(1), energy- and resource-efficient designs are likely to be favoured for support.

Common misconceptions

  • "CADA creates an AI Factory Fund." No. It leverages existing and planned programmes (Digital Europe, Horizon Europe, InvestEU, the future ECF) and encourages national and private investment.
  • "AI factories are only for big tech." The proposal explicitly aims for broad access, including SMEs, researchers and public bodies.
  • "An AI factory is just a bigger data centre." It is a type of facility optimised for AI workloads and treated as a strategic asset under Article 7(2)(e), not merely large-scale hosting.
  • "Funding is automatic." Support is competitive and conditional — aligned with the Leadership Initiatives' objectives, often requiring co-funding, and subject to State aid rules.

Related

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