Summary The Draghi report on European competitiveness serves as the strategic blueprint for the proposed Cloud and AI Development Act (CADA), specifically framing frontier AI as a critical "strategic asset" essential for technological sovereignty. As proposed, CADA operationalizes this by establishing strict criteria for "frontier AI priority projects" under Article 8 and mandating a resource-pooling mechanism under Article 9. This approach requires projects to be collaborative (involving at least three Member States) and to pool computing time, ensuring that the EU leverages collective capacity to compete globally rather than relying on fragmented national efforts or third-country providers.

Detail

The Cloud and AI Development Act (CADA) is explicitly designed as the legislative response to the strategic imperatives outlined in Mario Draghi's report, The Future of European Competitiveness. The report identifies a critical vulnerability: the EU's limited data centre capacity and its pronounced dependence on a limited pool of third-country providers, with non-EU hyperscalers controlling over 70% of the European cloud market. Draghi argues that computing infrastructures are no longer mere technical assets but have become "strategic resources critical to the Union's economic security, sovereignty, resilience, and competitiveness."

CADA translates these high-level economic security goals into specific legal mechanisms. The explanatory memorandum states that the proposal aims to address the EU's limited data centre capacity and dependence on third-country providers. It emphasizes that frontier AI technologies are advancing rapidly and have become "critical strategic assets." Consequently, strengthening the Union's capacity to develop and govern them is essential to ensure the AI transition aligns with Union values, safety standards, and long-term economic interests. The report's call to "regain and retain control over data and cloud computing services" is directly mirrored in CADA's focus on reducing dependencies and fostering homegrown capabilities.

The Draghi Report's Influence on Strategic Autonomy

The Draghi report identifies the current landscape as characterized by a dependence that exposes European users to operational discontinuity risks and subjects them to third-country jurisdictions with extraterritorial laws that may conflict with EU fundamental rights. In response, CADA establishes a framework to mitigate these risks by reinforcing the Union's capacity to develop and deploy cloud and AI technologies within its territory.

Recital 16 of the proposal notes that frontier AI technologies are advancing rapidly and are expected to have a profound impact on the Union's economy and society. As those technologies have become "critical strategic assets," strengthening the Union's capacity to develop and govern them is essential. The proposal explicitly states that "fostering the development of frontier AI technologies as strategic assets should reduce current dependencies on third-country technologies and strengthen the Union's AI ecosystem."

This strategic autonomy context is central to the definition of "frontier AI" in Article 2(4) of CADA, which defines it as "AI models or AI systems built upon such models that can perform a wide variety of tasks and that approach, reach or exceed the current state of the art." The legislation treats these models not just as commercial products, but as assets requiring sovereign control to ensure the EU's long-term competitiveness.

Article 8: Criteria for Frontier AI Priority Projects

To operationalize the Draghi report's call for targeted actions, CADA introduces Article 8, which sets out the criteria for the Commission to recognize projects as "frontier AI priority projects." These projects are selected through open calls for expression of interest and must support "grand challenge 3" (Frontier AI) as set out in Annex I.

The criteria under Article 8 are designed to ensure that supported projects are not isolated national efforts but collaborative, Union-wide initiatives. A project must meet three cumulative conditions:

  1. Pioneering Nature: It must be a pioneering project focused on the support and scaling-up of frontier AI technologies.
  2. Collaborative Structure: It must be undertaken by a European digital infrastructure consortium (EDIC) established pursuant to Decision (EU) 2022/2481 or another legal entity eligible for funding under Union law, and it must involve the participation of at least three Member States.
  3. Resource Pooling: The participating Member States must pool computing time and other relevant resources to support the implementation of the designated project.

This structure directly addresses the Draghi report's emphasis on collective effort. By requiring multi-Member State participation and resource pooling, CADA ensures that frontier AI development is a shared European endeavor, leveraging collective resources to achieve a scale that individual Member States could not achieve alone. This prevents a "race to the bottom" and ensures that strategic investments are coordinated at the Union level.

Article 9: Computing Support and Resource Allocation

Article 9 of CADA provides the concrete mechanism for supporting these priority projects through compute capacity. The Draghi report highlights that computing infrastructures are critical to economic security. CADA responds by mandating that the Union and Member States ensure sufficient AI computing resources from their compute capacities are allocated to support frontier AI priority projects.

Specifically, Article 9(2) states that 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 (EuroHPC) access time." This matching mechanism creates a powerful incentive for Member States to contribute resources, knowing the Union will amplify their investment.

Furthermore, Article 9(3) requires the Union and Member States to "endeavour to provide sufficient computing resource for AI industrial innovation, physical AI and public sector AI projects." This ensures that frontier AI development is not siloed but contributes to broader industrial and public sector capabilities, aligning with the Draghi report's goal of strengthening homegrown cloud and AI capabilities across the entire value chain.

Scaling and Resource-Pooling Rationale

The rationale behind Articles 8 and 9 is rooted in the economic reality of frontier AI development, which is capital-intensive and requires unprecedented scale. The explanatory memorandum notes that the emergence of frontier AI requires a collaborative approach at the Union level due to their technical complexity and capital-intensive nature.

By requiring the pooling of computing time and resources (Article 8(c)), CADA ensures that the significant costs of training and scaling frontier models are distributed across multiple Member States. This reduces the financial burden on any single country while creating a unified front in the global AI race. The Commission's role in recognizing these projects (Article 8) and matching resources (Article 9) provides the necessary coordination and incentive structure to make this pooling effective.

This approach also supports the broader CADA objective of increasing computing capacity deployed in the EU through innovative and sustainable technologies. By directing EuroHPC resources toward frontier AI priority projects, CADA ensures that the EU's existing high-performance computing infrastructure is used strategically to develop sovereign AI capabilities, rather than being underutilized or dependent on external providers. The proposal aims to "bridge the gap between the Union's advanced research and innovation capabilities and their sustainable exploitation," a direct echo of Draghi's call for a robust financial and talent flywheel.

What this means for you

For public-sector procurement officers, national strategy planners, and research consortium leaders, the Draghi report's influence on CADA's frontier AI provisions implies a fundamental shift in how AI projects are identified, structured, and supported.

  1. Prioritize Collaborative Consortia: When evaluating or proposing AI initiatives, the focus must shift from national silos to cross-border collaboration. Article 8 explicitly requires participation from at least three Member States for frontier AI priority projects. Procurement strategies and grant applications should favor consortia that demonstrate this collaborative structure, as they are the only entities eligible for Commission recognition and enhanced compute support.
  2. Align with Grand Challenges: Ensure that proposed projects align with the "grand challenges" outlined in CADA Annex I, particularly Grand Challenge 3 (Frontier AI). Projects must demonstrate how they contribute to scaling up frontier AI technologies as strategic assets, moving beyond incremental improvements to pioneering breakthroughs.
  3. Leverage Compute Pooling: Actively participate in pooling computing time and resources with other Member States. Article 9 mandates that the Union will match resources contributed by Member States to recognized priority projects. By contributing to these pooled resources, your institution can access additional EuroHPC capacity, effectively amplifying your project's computational power and reducing individual costs.
  4. Focus on Sovereignty and Security: Given the Draghi report's emphasis on reducing third-country dependencies, ensure that proposed frontier AI projects incorporate measures to safeguard data sovereignty and operational autonomy. This aligns with CADA's broader sovereignty framework and increases the likelihood of project recognition as a "priority."
  5. Integrate with National Strategies: Article 7 of CADA requires Member States to adopt national cloud and AI strategies. Ensure that your procurement activities and project proposals are consistent with these national strategies, which should in turn be aligned with the objectives of CADA and the Draghi report's competitiveness goals.

Common misconceptions

Misconception 1: CADA funds all frontier AI research. CADA does not provide blanket funding for all frontier AI research. Instead, it establishes a framework for recognizing specific "frontier AI priority projects" that meet strict criteria under Article 8, including multi-Member State collaboration and resource pooling. Support is targeted and conditional on meeting these strategic objectives.

Misconception 2: The Draghi report is legally binding. The Draghi report is a strategic document that influences policy but is not itself legislation. CADA is the legislative instrument that translates the report's recommendations into binding rules. The reference to the Draghi report in CADA's recitals and explanatory memorandum provides context and justification for the legal obligations, but the enforceable requirements are found in the articles of the regulation itself.

Misconception 3: Frontier AI projects can be national solo efforts. Article 8 explicitly requires that frontier AI priority projects involve the participation of at least three Member States and be undertaken by a European digital infrastructure consortium or similar eligible entity. Solo national projects do not meet the criteria for recognition as priority projects under CADA.

Misconception 4: Compute support is guaranteed for all AI projects. Article 9 specifies that compute support is allocated to "frontier AI priority projects" that have been recognized under Article 8. Additionally, the Union's matching of resources is subject to the availability of sufficient AI computing capacity within the Union's share of EuroHPC access time. Support is not automatic for all AI initiatives.

Official sources

Related

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