Summary Grand Challenge 3 under the proposed Cloud and AI Development Act (CADA) is dedicated to "Developing the next generation of multimodal frontier AI models and systems and pioneering novel capabilities." As defined in Annex I of the proposal, this initiative targets superior performance in advanced reasoning, cross-modal understanding, and agentic capabilities. To qualify for specific Union support as a "frontier AI priority project" under Article 8, an initiative must directly support this grand challenge, be undertaken by a European digital infrastructure consortium (EDIC) or eligible entity, involve at least three Member States, and ensure those states pool computing resources. Applications highlighted include scientific discovery, world models, and automated management simulation.

Detail

The Cloud and AI Development Act (CADA), as proposed in COM(2026) 502 final, establishes a comprehensive framework to strengthen Europe's cloud and AI ecosystem. A central pillar of this framework is the "Cloud and AI Leadership Initiatives," designed to foster research, innovation, and the deployment of large-scale capacity. These initiatives are structured around eight strategic "Grand Challenges" detailed in Annex I of the regulation. Grand Challenge 3 is exclusively focused on Frontier AI.

Defining Grand Challenge 3: Scope and Objectives

According to Annex I of the CADA proposal, Grand Challenge 3 is titled "Frontier AI." Its explicit objective is "Developing the next generation of multimodal frontier AI models and systems and pioneering novel capabilities."

The proposal delineates the technical and functional scope of this challenge with precision, moving beyond simple scaling to focus on architectural and cognitive breakthroughs:

  1. Architectural Design and Development: The core focus is on creating next-generation multimodal models and systems that "push the boundaries of current algorithmic capabilities." The proposal specifically identifies three key areas of superior performance:

    • Advanced Reasoning: The ability of AI systems to perform complex logical deductions, solve multi-step problems, and navigate abstract concepts.
    • Cross-modal Understanding: The capacity to seamlessly interpret, integrate, and reason across diverse data types, such as text, images, audio, and video, within a single model architecture.
    • Agentic Capabilities: The development of AI systems capable of autonomous action. These systems, often termed "AI agents," can perceive their environment, use tools, and adapt to changing inputs to achieve specific goals without constant human intervention.

  2. Novel Approaches to Efficiency and Structure: Grand Challenge 3 encourages the investigation of "novel approaches to model efficiency, cognitive modelling, and alternative computational structures." This signals a strategic intent to explore architectures beyond current dominant paradigms (such as standard transformer models) if they offer superior efficiency or capability. The goal is to achieve breakthroughs in how AI processes information, not just in the volume of data it consumes.

  3. Strategic Applications: The proposal outlines specific high-impact application domains where these frontier capabilities would be deployed:

    • Foundational Science: Leveraging AI for "scientific discovery and complex data interpretation." This includes accelerating research in fields like biology, physics, and chemistry by enabling AI to analyze vast datasets and generate hypotheses.
    • World Models: Developing models capable of creating internal representations of the world to "improve reasoning, automated management simulation and planning." These models would allow for the simulation of complex real-world scenarios to test strategies before implementation.
    • Automated Management Simulation: Using AI to simulate complex industrial or societal systems to enhance decision-making processes.

The Pathway to Priority Status: Article 8 and Frontier AI Priority Projects

Grand Challenge 3 is not merely a research theme; it serves as the gateway to specific Union-level support mechanisms. Article 8 of the CADA proposal establishes the criteria for designating a project as a "frontier AI priority project."

To receive this designation, a project must be selected through an open call for expression of interest and must explicitly support Grand Challenge 3 as set out in Annex I. The proposal mandates three cumulative criteria for such projects:

  1. Pioneering Focus: The project must be "pioneering" and specifically focused on the "support and scaling-up of frontier AI technologies." This ensures that only projects at the cutting edge of the field qualify.
  2. Eligible Entity and Cross-Border Participation: The project 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." Crucially, it must involve the participation of at least three Member States. This requirement enforces a pan-European approach, preventing fragmentation and ensuring broad strategic alignment.
  3. Resource Pooling: The participating Member States must "pool computing time and other relevant resources to support the implementation of the designated project." This criterion ensures that the massive computational demands of frontier AI are met through shared European capacity, reinforcing the Act's broader goals of sovereignty and capacity building.

This linkage ensures that Grand Challenge 3 is pursued through collaborative, cross-border European efforts that leverage shared computational resources, rather than in isolation by single entities.

Strategic Context and Rationale

The emphasis on Grand Challenge 3 reflects the EU's strategic imperative to reduce dependence on non-European providers for critical AI technologies. The explanatory memorandum notes that "frontier AI technologies are advancing rapidly and are expected to have a profound impact on the Union's economy and society." As these technologies become "critical strategic assets," strengthening the Union's capacity to develop and govern them is essential.

The proposal argues that fostering the development of frontier AI as strategic assets is necessary to "reduce current dependencies on third-country technologies and strengthen the Union's AI ecosystem." Furthermore, it aims to ensure that the AI transition "is aligned with Union values, safety standards and long-term economic interests." By targeting advanced reasoning and agentic capabilities, the EU seeks to foster an ecosystem capable of competing globally in the most advanced segments of the AI market while maintaining control over the underlying technologies.

What this means for you

For technology leaders, researchers, and public sector bodies, Grand Challenge 3 presents a distinct set of opportunities and strategic imperatives under the proposed CADA.

For AI Researchers and Developers

If your organization is pioneering next-generation AI models, particularly those involving multimodal inputs or agentic behaviors, Grand Challenge 3 offers a pathway to Union support. However, the Article 8 criteria impose a strict structural requirement: you cannot apply alone. You must collaborate with partners from at least two other Member States. This necessitates the formation of consortia, potentially through European digital infrastructure consortiums (EDICs). The focus on "novel approaches" and "alternative computational structures" suggests that innovation in efficiency and architecture is as valued as raw model scale.

For SMEs and Start-ups

While frontier AI is computationally intensive, the CADA proposal explicitly recognizes the role of SMEs and small mid-caps (SMCs). You may not lead a Grand Challenge 3 project, but you can contribute specialized components, such as novel algorithms for efficiency, domain-specific data for scientific discovery, or tools for cognitive modelling. The proposal encourages engagement with the network of Experience and Acceleration Centres for AI (Centres for AI) established under Article 5. These centres are tasked with supporting SMEs in their digital transformation and connecting them with European providers of cloud and AI technologies.

For Public Sector and Research Institutions

Public sector bodies and research institutions are primary beneficiaries of the applications highlighted in Grand Challenge 3. As the EU invests in scientific discovery and world models, expect increased availability of sovereign, high-performance AI tools for research and complex data interpretation. This could significantly accelerate local research agendas and improve public service delivery through better simulation and planning tools, particularly in critical domains like healthcare and crisis management.

For Infrastructure Providers

The requirement for Member States to "pool computing time" for frontier AI priority projects (Article 8(1)(c)) underscores the critical importance of accessible, high-capacity compute infrastructure. Providers of cloud and High-Performance Computing (HPC) services should prepare for increased demand from these collaborative projects. Ensuring that your infrastructure can support the rigorous requirements of frontier AI training and inference will be essential for participating in the ecosystem.

Common misconceptions

Misconception 1: Grand Challenge 3 is only for large tech giants. While frontier AI requires significant resources, the CADA proposal explicitly encourages collaboration and the involvement of diverse entities. The requirement for cross-border participation (at least three Member States) and the role of European digital infrastructure consortiums mean that smaller players, including SMEs and academic institutions, can participate as partners in these large-scale initiatives.

Misconception 2: Any advanced AI project qualifies as a frontier AI priority project. Not all advanced AI projects qualify. To be designated as a frontier AI priority project under Article 8, a project must specifically support Grand Challenge 3 as defined in Annex I. It must also meet strict criteria regarding the legal entity structure (e.g., EDIC), cross-border participation, and resource pooling. Projects focused on other grand challenges (e.g., Industrial AI or Physical AI) do not qualify for this specific designation.

Misconception 3: Grand Challenge 3 focuses solely on model size. The proposal emphasizes "novel capabilities" such as advanced reasoning, cross-modal understanding, and agentic behaviors, rather than just scaling up model parameters. It also encourages investigating "novel approaches to model efficiency, cognitive modelling, and alternative computational structures." This means that innovation in architecture and efficiency is just as important as raw scale.

Misconception 4: The EU is restricting the development of frontier AI. On the contrary, Grand Challenge 3 is designed to boost the development of frontier AI within the EU. The goal is to reduce dependence on non-European providers and ensure that the EU remains competitive in this strategic area. The framework aims to create an enabling environment for innovation while ensuring that the development of these powerful technologies aligns with EU values and safety standards.

Official sources

Related

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