Summary The proposed Cloud and AI Development Act (CADA) directly targets the automotive industry as a strategic sector for industrial and physical AI. As proposed, CADA would fund research into software-defined vehicles and autonomous driving, facilitate access to high-performance computing for testing, and establish a sovereign cloud framework that could influence supply chains serving public authorities. For automotive CTOs and SMEs, this means new funding opportunities for frontier AI projects, access to national "Centres for AI" for testing, and a potential requirement to use EU-based cloud services when supplying public-sector clients.

Detail

The Cloud and AI Development Act (CADA), proposed by the European Commission in June 2026 (COM(2026) 502 final), is designed to strengthen Europe's cloud and AI ecosystem. While it is a broad framework, it contains specific provisions that explicitly target the automotive industry, recognizing it as a key strategic sector for industrial AI and physical AI development. The impact on the automotive sector is threefold: direct research and innovation support, infrastructure acceleration for data centres, and sovereignty requirements for cloud services that may ripple through the supply chain.

Direct Support for Automotive AI and Software-Defined Vehicles

CADA explicitly identifies the automotive sector as a priority for AI adoption. Recital 19 of the proposal states that advancements in AI should "support the development, testing and deployment of innovative software platforms contributing to the Union industrial leadership in software defined vehicles and autonomous driving." This recital highlights a strategic intent to reduce obstacles in testing and deploying AI models, particularly within cities and regions, to enable the safe and trustworthy deployment of connected and autonomous mobility solutions.

To achieve this, CADA establishes the Cloud and AI Leadership Initiatives. Under Article 4, operational objective 5 focuses on accelerating the development and uptake of industrial AI across strategic sectors, including automotive. This involves:

  • Accelerating the development of sectoral AI models tailored to industrial needs.
  • Facilitating access to necessary computing resources and AI tools.
  • Enabling secure large-scale data pooling for collaborative AI training while preserving confidentiality.

Furthermore, Article 8 allows the Commission to recognize "frontier AI priority projects." These are pioneering projects that support Grand Challenge 3 (Frontier AI) as set out in Annex I. While Grand Challenge 5 (Industrial AI) and Grand Challenge 4 (Physical AI) are critical for the automotive sector, the specific "frontier AI priority project" designation under Article 8 is legally limited to projects supporting Grand Challenge 3. However, automotive projects focusing on industrial or physical AI would still benefit from the general operational objectives under Article 4 and the broader support mechanisms of the Leadership Initiatives. For the automotive industry, this means that large-scale projects involving autonomous driving, robotics, or advanced vehicle-to-everything (V2X) communication could qualify for special status and support, provided they align with the specific criteria of the designated grand challenges.

Computing Resources and Testing Infrastructure

A major bottleneck for automotive AI is access to high-performance computing (HPC) for training complex models. Article 9 of CADA addresses this by ensuring that sufficient AI computing resources from Union and Member State capacities are allocated to support frontier AI priority projects. The Union would match AI computing resources contributed by Member States to these designated projects, within the limits of available EuroHPC capacity.

Additionally, CADA mandates the creation of Experience and Acceleration Centres for AI (Centres for AI) under Article 5. These centres, built on existing European Digital Innovation Hubs, would support the integration and scaling-up of AI use cases in strategic industrial sectors. For automotive SMEs and CTOs, these centres would serve as entry points to the European AI innovation ecosystem, providing access to testing facilities, upskilling schemes, and connections with European cloud and AI providers. Recital 19 specifically notes that Member States should facilitate the development, testing, and deployment of AI systems for autonomous driving through cooperation with these Centres for AI, the automotive industry, suppliers, cities, and regions.

Sovereign Cloud and Data Centre Acceleration

CADA introduces a Union cloud computing sovereignty framework with four assurance levels (Article 16). While this primarily targets public sector procurement, it has downstream effects on the automotive industry. Automotive manufacturers often supply software and data processing services to public bodies (e.g., for smart city infrastructure, public transport, or traffic management). To win these contracts, suppliers may need to use cloud services recognized at specific Union assurance levels (2, 3, or 4 for high-risk public order activities).

The sovereignty criteria in Annex II require that for higher assurance levels, infrastructure, assets, and personnel must be located in the Union, and customer data must remain exclusively within the Union.

  • Level 2: Requires infrastructure and assets in the Union. Personnel requirements are conditional: "if the public sector body determines that imposing additional personnel screening and Union citizenship requirements are necessary, the audited provider should ensure that personnel meeting those requirements are available" (Annex II, 2.1(d)).
  • Level 3 & 4: Require that personnel involved in the provision of the service "are Union citizens" (Annex II, 3.1(d) and 4.1(d)). For Level 3 and 4, there is also a strict prohibition on third-country control, though Level 3 allows a derogation if the Commission adopts an implementing act under Article 18 identifying a third country with sufficient safeguards. Level 4 strictly prohibits third-country control (Annex II, 4.1(g)).

This pushes automotive software platforms toward EU-based infrastructure to ensure compliance when serving public sector clients.

Simultaneously, CADA aims to accelerate the deployment of data centres through data centre acceleration zones (Article 10). These zones offer streamlined permitting (max 12 months, per Article 13) and single information points (Article 12) for data centre projects. For automotive companies relying on low-latency edge computing for autonomous driving, the rapid deployment of these sustainable, high-capacity data centres is critical. The proposal aims to triple EU data centre capacity in five to seven years, directly supporting the compute-intensive needs of the automotive sector.

Procurement and Innovation

Article 32 introduces "Union added value" criteria for public procurement of cloud services and AI systems. Contracting authorities must evaluate tenders based on their contribution to strengthening the European digital supply chain, including the use of hardware and software designed or manufactured in the Union. Automotive companies bidding for public sector AI or cloud contracts will need to demonstrate their alignment with these European value chain criteria.

What this means for you

For CTOs, architects, and SMEs in the automotive sector, CADA presents both opportunities and compliance considerations:

  1. Access to Funding and Compute: If your company is developing frontier AI for autonomous driving or industrial AI, monitor calls for "frontier AI priority projects" under Article 8. Qualifying could unlock matched AI computing resources from the Union's EuroHPC capacity. Note that Article 8 specifically targets Grand Challenge 3, but other automotive AI projects may still qualify under the broader operational objectives of Article 4.
  2. Leverage Centres for AI: Engage with the national Centres for AI established under Article 5. These hubs will provide testing environments, skills training, and connections to European cloud providers, reducing the barrier to entry for SMEs adopting AI.
  3. Sovereign Cloud Readiness: If you supply software or data services to public authorities, ensure your cloud infrastructure can meet the Union assurance levels defined in Annex II. This may require shifting data processing and storage to EU-based providers. Be aware that for Levels 3 and 4, personnel handling the service must be Union citizens, whereas for Level 2, this is conditional on the public body's requirements.
  4. Public Procurement Strategy: When bidding for public sector contracts involving AI or cloud services, highlight your "Union added value" (Article 32). Emphasize the use of EU-designed hardware, software, and research outcomes to strengthen the European supply chain.
  5. Data Centre Planning: For companies needing low-latency edge computing, track the designation of data centre acceleration zones in your region. These zones offer faster permitting (max 12 months) and better grid connectivity, crucial for deploying the infrastructure needed for real-time autonomous driving systems.

Common misconceptions

  • Misconception: CADA only affects large tech companies.
    • Reality: CADA explicitly supports SMEs and start-ups. Article 33(4) sets an objective for Member States to award at least 25% of their procurement for cloud computing services and AI systems to innovative SMEs. The Centres for AI are also designed to support SMEs in digital transformation.
  • Misconception: Sovereign cloud requirements apply to all private sector data.
    • Reality: The mandatory use of Union assurance levels applies primarily to public sector bodies and Union entities (Article 30). However, private entities in critical sectors (like automotive) may voluntarily conduct impact assessments (Article 31) and may face indirect pressure if they supply public sector clients who require sovereign cloud services.
  • Misconception: CADA replaces the AI Act.
    • Reality: CADA complements the AI Act. The AI Act regulates the safety and fundamental rights implications of AI systems, while CADA focuses on building the underlying infrastructure, computing capacity, and sovereignty framework to support AI development and deployment.
  • Misconception: Data must never leave the EU for any cloud service.
    • Reality: Under Union assurance level 1, data can leave the Union if the public sector body explicitly requires otherwise. Higher assurance levels (2-4) have stricter data localization requirements, but they are tiered based on risk assessments.
  • Misconception: All automotive AI projects qualify for "frontier AI priority" status.
    • Reality: While automotive projects are strategic, the specific "frontier AI priority project" designation under Article 8 is limited to projects supporting Grand Challenge 3 (Frontier AI). Projects focused solely on Industrial AI (Challenge 5) or Physical AI (Challenge 4) would fall under general operational objectives but not the specific Article 8 designation mechanism unless they also address Challenge 3.

Official sources

Related

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