Summary Under the proposed Cloud and AI Development Act (CADA), the "AI first" principle is a strategic mandate requiring organisations to proactively reflect on their business processes to identify opportunities offered by AI, while simultaneously weighing potential risks. As proposed, this principle is not optional; it must be embedded in Member States' national cloud and AI strategies (Article 7) and serves as a core operational objective for the new Experience and Acceleration Centres for AI (Article 5). The principle aims to shift the EU's digital transformation from reactive technology adoption to proactive, AI-driven innovation, supported by the broader Cloud and AI Leadership Initiatives.

Detail

The Cloud and AI Development Act (CADA), as proposed in COM(2026) 502 final, is a legislative instrument designed to strengthen Europe's cloud and AI ecosystem, reduce critical dependencies on non-European providers, and safeguard the Union's technological sovereignty. A central pillar of this proposal is the acceleration of AI adoption across the public and private sectors. To achieve this, CADA codifies the "AI first" principle as a foundational concept for national planning and local implementation.

Defining the "AI first" principle

The "AI first" principle is not a mandate to deploy AI in every conceivable scenario, nor is it a directive to replace human decision-making indiscriminately. Instead, it is a strategic directive for how organisations should approach the design of their operations.

As explicitly defined in the proposal, the principle urges organisations to "reflect on their business processes, considering the needs and opportunities offered by AI, while taking into consideration the potential risks." This definition, found in the context of national strategies, establishes a dual obligation:

  1. Proactive Consideration: Organisations must actively seek out where AI can transform workflows, improve service delivery, or solve complex problems. AI is to be considered as a primary tool from the outset of a project, rather than an afterthought or a legacy system replacement added at the end of a lifecycle.
  2. Risk Awareness: This proactive stance is balanced by a mandatory consideration of risks. The principle does not encourage reckless adoption; it requires that opportunities be weighed against potential harms, ensuring that any AI integration is safe, ethical, and compliant with existing frameworks like the AI Act and GDPR.

In practice, an "AI first" approach means that when a public authority designs a new citizen service or a business streamlines a supply chain, the initial question is "How can AI improve this?" rather than "How can we digitise this manually?"

Where the principle appears in CADA

The "AI first" principle is explicitly codified in two critical areas of the proposal, creating a top-down and bottom-up framework for adoption.

1. National Cloud and AI Strategies (Article 7) Article 7 imposes a binding obligation on Member States to establish national cloud and AI strategies within one year of the Regulation's entry into force. Article 7(2)(a) specifically mandates that these strategies must include "key objectives and priorities for cloud and AI adoption, in line with the 'AI first' principle."

This provision ensures that every Member State aligns its national digital agenda with this proactive approach. It creates a unified baseline where national governments must explicitly prioritise AI adoption in their strategic planning, moving beyond general digitalisation goals to specific AI-driven objectives. The strategies must also include governance frameworks to achieve these priorities, ensuring that the "AI first" mindset is translated into actionable policy.

2. Experience and Acceleration Centres for AI (Article 5) Article 5 establishes a network of Experience and Acceleration Centres for AI (building on the existing European Digital Innovation Hubs). These Centres are tasked with supporting the integration and scaling of AI use cases across the Union. Article 5(2)(b) states that the objectives of these Centres include accelerating the broad adoption of cloud and AI technologies at regional and local levels, "notably for SMEs, SMCs and public sector bodies, in line with the 'AI first' principle."

This operationalises the principle at the local level. The Centres are not merely information hubs; they are active accelerators mandated to promote this specific mindset. They are required to help organisations reflect on their processes and identify AI opportunities, ensuring that the "AI first" principle reaches the grassroots level of the economy, including small and medium-sized enterprises (SMEs) and small mid-caps (SMCs).

Connection to the Cloud and AI Leadership Initiatives

The "AI first" principle is the demand-side engine that drives the supply-side innovations of the Cloud and AI Leadership Initiatives (Title II of CADA). These Initiatives, established under Article 3, aim to bridge the gap between advanced research and sustainable exploitation by supporting the development of cutting-edge technologies and large-scale capacity.

The operational objectives of these Initiatives, detailed in Article 4, directly reinforce the "AI first" mandate:

  • Operational Objective 5 focuses on "accelerating the development and uptake of industrial AI across the Union's strategic sectors."
  • Operational Objective 7 targets "increasing the development and adoption of AI models and systems across the Union's public sectors."
  • Operational Objective 8 aims to "promote the broad adoption of AI by private and public sector organisations... through the network of Experience and Acceleration Centres for AI."

By mandating the "AI first" principle in national strategies, CADA creates a coordinated demand for the technologies and infrastructure developed under the Leadership Initiatives. It ensures that as the EU builds sovereign AI capabilities, there is a parallel, strategic effort across all Member States to integrate these tools into public services and industrial processes. The Initiatives provide the "fuel" (technology and capacity), while the "AI first" principle ensures the "engine" (organisations) is ready to use it.

What this means for you

For public-sector procurement officers, digital transformation leads, and business strategists, the "AI first" principle has immediate and long-term implications for planning and decision-making.

  • Strategic Alignment: Your national government is required to produce a national cloud and AI strategy that explicitly adopts the "AI first" principle. As a public official or business leader, you must align your department's or company's digital roadmap with this national strategy. This means justifying digital investments not just on cost or immediate utility, but on how they enable AI-driven efficiency and innovation.
  • Procurement and Design: When drafting tenders or designing new services, you should consider whether AI solutions can deliver better outcomes than traditional software. The "AI first" mindset encourages you to ask: "How can AI improve this service?" before settling on a non-AI alternative. This aligns with CADA's goal of fostering the uptake of AI across the public sector.
  • Risk Management: The principle explicitly mentions "taking into consideration the potential risks." This means you cannot adopt AI blindly. You must conduct thorough risk assessments, ensuring compliance with the AI Act (for high-risk systems), GDPR (for data protection), and the new sovereignty requirements in CADA (such as Union Assurance Levels). An "AI first" approach includes "security and sovereignty first."
  • Utilising Support Structures: You are encouraged to engage with your local Centre for AI. These centres, mandated by CADA to operate under the "AI first" principle, offer expertise, testing facilities, and skills support. They can help you identify where AI can be safely integrated into your services and provide the technical assistance needed to procure and implement these solutions effectively.

Common misconceptions

"AI first means AI always." This is incorrect. The principle does not mandate that every single administrative task must be automated by AI. It requires that AI be considered as a primary option during the design phase. If a simple rule-based script, a human-led process, or a non-AI digital solution is more appropriate, secure, or cost-effective, that is a valid outcome of an "AI first" assessment. The principle is about informed consideration, not forced adoption.

"It overrides existing laws like the AI Act." No. CADA complements the AI Act. The "AI first" principle must be applied "while taking into consideration the potential risks." This means that any AI adoption must still comply with the strict safety, transparency, and fundamental rights requirements of the AI Act. You cannot use "AI first" as an excuse to bypass compliance checks or deploy high-risk AI systems without proper assessment.

"It only applies to large government departments." False. Article 5 explicitly mentions SMEs, small mid-caps (SMCs), and local public sector bodies. The "AI first" principle is designed to be scalable. For smaller entities, it might mean using pre-built AI tools from the EuroCloud Federation or leveraging the support of Centres for AI to enhance efficiency, rather than building custom models.

"It is only about technology." The principle is also about process and skills. Reflecting on business processes involves organisational change. An "AI first" strategy requires upskilling staff, changing workflows, and rethinking service delivery models, not just buying new software.

Official sources

Related

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