Summary Under the proposed Cloud and AI Development Act (CADA), operational objective 7 is dedicated to increasing the development and adoption of AI models and systems across the Union's public sectors. As explicitly set out in Article 4(7) of the proposal, this objective mandates accelerating technological uptake in critical domains, improving public service delivery and decision-making, promoting the sharing and reuse of training data and AI models, and facilitating secure, privacy-enhancing health data reuse. It also specifically supports the development and deployment of AI in the automotive sector, including for autonomous driving. This objective operates in tandem with Grand Challenge 8 ("Public Sector AI") to build a resilient, sovereign, and high-impact public sector AI ecosystem.

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 support research, innovation, and large-scale capacity building. Within these initiatives, Article 4 of the proposal outlines specific operational objectives. Operational objective 7 is dedicated exclusively to the public sector, recognizing that public authorities are both major consumers of AI and key drivers of societal benefit.

Legal Basis and Core Requirements

Article 4(7) of the CADA proposal explicitly lists the measures the Cloud and AI Leadership Initiatives shall pursue under operational objective 7. These measures are cumulative and designed to transform how public services are delivered and how public data is utilized. The provision states that the initiatives shall:

  1. Accelerate technological development and uptake in critical domains: The proposal aims to speed up the deployment of AI models and systems in sectors deemed critical for public interest. This ensures that high-impact areas receive prioritized support for AI integration, addressing the urgent need for modernization in essential services.
  2. Improve service delivery and decision-making: A core goal is to develop AI tools that increase the effectiveness of public service delivery and accessibility for the general public. This includes simplifying administrative procedures and supporting better, data-driven decision-making within public administrations, thereby reducing unnecessary burdens on citizens and businesses.
  3. Promote sharing and reuse: To avoid fragmentation and duplication of effort, the initiative promotes the sharing and reusing of training data and AI models across the Union's public services. This interoperability allows successful solutions to be scaled up and adapted across different Member States and administrative levels, fostering a unified European digital public space.
  4. Facilitate health data reuse: Recognizing the sensitivity and immense value of health data, the proposal specifically facilitates "secure, privacy-enhancing health data reuse for AI models and tools in healthcare." This aims to improve clinical decisions and transform the pharmaceutical sector while maintaining strict security and data protection standards, ensuring that patient privacy is not compromised in the pursuit of innovation.
  5. Support automotive and autonomous driving: The objective extends to the automotive sector, facilitating the development, testing, and deployment of AI models and tools, including those for autonomous driving. This involves reducing obstacles to test and deploy AI models, particularly within cities and regions, to contribute to Union leadership in software-defined vehicles and autonomous mobility solutions.

Link to Grand Challenge 8

Operational objective 7 is closely aligned with Grand Challenge 8, as outlined in Annex I of the CADA proposal. Grand Challenge 8 is titled "Public Sector AI" and focuses on developing AI models and systems based on high-quality data from the public sector targeting critical domains such as healthcare, public administration, law, and crisis management.

While operational objective 7 sets the strategic direction for adoption and deployment, Grand Challenge 8 provides the technical and research roadmap for developing these specific models. The Grand Challenge explicitly targets "public service solutions that are expected to have a high positive impact on the most critical public services and are shared across different levels of public sector organisations." It further emphasizes the use of "privacy-preserving frameworks, (such as federated learning and high-fidelity synthetic data generation)" to train models without compromising the confidentiality of underlying datasets. Together, these provisions ensure that public sector AI is not only adopted but is also built on robust, high-quality, and privacy-preserving foundations.

Strategic Context

The emphasis on public sector AI is driven by the need to reduce dependencies on non-European cloud and AI providers. By fostering a domestic ecosystem for public sector AI, the EU aims to ensure that critical infrastructure and sensitive data remain under European control. This aligns with the broader CADA goals of technological sovereignty and resilience. The proposal also encourages the use of open standards and open-source components in building these public sector AI stacks, further enhancing transparency and reducing vendor lock-in. The objective is to create a "European public sector cloud federation" (EuroCloud Federation) that facilitates the sharing of secure and resilient public-sector data centre services and cloud computing services, thereby enabling the scaling of these AI solutions.

What this means for you

For public-sector bodies, procurement officers, and AI developers, operational objective 7 under CADA signals a significant shift in how AI projects should be conceived, funded, and procured.

Prioritize Critical Domains and Service Improvement When planning AI initiatives, prioritize projects that directly impact critical public domains such as healthcare, crisis management, public administration, and law. Procurement strategies should favor solutions that demonstrably improve service delivery, accessibility, and decision-making. Look for vendors who can show how their AI systems simplify administrative burdens and enhance citizen access to services, aligning with the mandate to "simplify administrative procedures."

Leverage Data Sharing and Reuse Avoid building siloed AI solutions. Instead, seek opportunities to share and reuse training data and AI models across different departments and Member States. This not only reduces costs but also accelerates deployment by building on existing, validated solutions. Ensure that your data governance frameworks support secure, privacy-enhancing data sharing, particularly in sensitive areas like health. For health-related AI, prioritize solutions that employ advanced privacy-preserving techniques, such as federated learning or synthetic data generation, to enable data reuse without compromising patient confidentiality.

Support Automotive and Autonomous Driving Innovation If your jurisdiction is involved in urban planning or transportation, consider how AI can support the development and testing of autonomous driving technologies. CADA encourages reducing obstacles to testing AI models in real-world environments, such as cities and regions. Procurement officers can facilitate this by creating sandboxes or testing grounds for AI-enabled connected and autonomous mobility solutions, in line with the objective to "facilitate the development, testing and deployment of AI models and tools in the automotive sector."

Align with Grand Challenge 8 When drafting tender specifications, align them with the goals of Grand Challenge 8. This means emphasizing high-quality data, privacy-preserving frameworks, and scalability. Look for solutions that can be shared across different levels of public sector organizations and that contribute to the broader European AI ecosystem. The proposal explicitly mentions that initiatives should rely on "specialised computing resources and testing facilities necessary to validate AI systems in real-world environments."

Focus on Sovereignty and Open Standards Given CADA's focus on sovereignty, prioritize AI solutions that are developed in the EU and comply with European data protection standards. Encourage the use of open standards and open-source components to ensure interoperability and reduce long-term dependencies on specific vendors. The proposal aims to "foster the development of innovative, competitive and resilient cloud and AI technologies" relying on open standards.

Common misconceptions

Misconception 1: Operational objective 7 only applies to large central governments. Correction: Article 4(7) and the broader CADA framework apply to public sectors at all levels, including regional and local authorities. The proposal explicitly mentions supporting AI adoption at regional and local levels, particularly through the network of Experience and Acceleration Centres for AI, which are tasked with accelerating broad adoption at these levels.

Misconception 2: Health data reuse under CADA means unrestricted access to patient data. Correction: The proposal specifically mentions "secure, privacy-enhancing health data reuse." This implies the use of advanced techniques such as federated learning or synthetic data generation that allow AI training without exposing raw, identifiable patient data. Strict adherence to GDPR and other data protection laws remains paramount, and the objective is to facilitate reuse "while ensuring security and data protection."

Misconception 3: Operational objective 7 replaces the AI Act's requirements for high-risk AI systems. Correction: CADA complements, but does not replace, the AI Act. Public sector AI systems that fall under the AI Act's definition of high-risk (e.g., in healthcare, law enforcement, or public administration) must still comply with the AI Act's strict requirements for risk management, data governance, and human oversight. CADA focuses on the supply-side and ecosystem-building aspects, while the AI Act focuses on safety and fundamental rights.

Misconception 4: The automotive focus is only for private sector car manufacturers. Correction: While car manufacturers are key players, CADA explicitly supports public sector involvement in facilitating the testing and deployment of AI models for autonomous driving. This includes public authorities providing testing environments, updating infrastructure, and supporting the regulatory frameworks needed for safe autonomous mobility, as well as Member States facilitating these efforts "in cooperation with the Centres for AI, the automotive industry, suppliers, cities and regions."

Official sources

Related

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