Summary Under the proposed Cloud and AI Development Act (CADA, COM(2026) 502 final — a draft regulation, not yet in force), Article 5(2) gives the Experience and Acceleration Centres for AI ("Centres for AI") three objectives: (1) support the integration and scaling-up of AI use cases in strategic industrial and public sectors; (2) accelerate the broad adoption of cloud and AI at regional and local level, notably for SMEs, small mid-caps (SMCs) and public sector bodies, in line with the "AI first" principle; and (3) leverage relevant infrastructure to accelerate the development and fine-tuning of AI models and systems. The specific tasks that deliver these objectives are listed separately in Article 5(3).

Detail

The Centres for AI are CADA's vehicle for turning AI research into broad, practical adoption. As proposed in Article 5, each Member State would establish them, building on the European Digital Innovation Hubs (EDIHs). Their three objectives are set out in Article 5(2).

1. Support the integration and scaling-up of AI use cases (Article 5(2)(a))

The first objective is to "support the integration and scaling-up of AI use cases in strategic industrial and public sectors." This is about moving organisations beyond isolated pilots to production-grade deployment, so that successful experiments become repeatable solutions delivering real value.

2. Accelerate broad adoption at regional and local level (Article 5(2)(b))

The second objective is to "accelerate 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 is the decentralisation objective: bringing expertise closer to smaller entities so they are not left behind.

The "AI first" principle is defined in the Apply AI Strategy and referred to in CADA's recitals as "urging organisations to reflect on their business processes, considering the needs of an opportunities offered by AI, while taking into consideration the potential risks." It is a principle of considered, proactive evaluation of where AI can add value — not a mandate to use AI everywhere.

3. Leverage infrastructure for model development (Article 5(2)(c))

The third objective is to "leverage relevant infrastructure to accelerate the development and fine-tuning of AI models and systems." This lowers the barrier for organisations that lack the in-house compute or data environments to train or adapt models.

From objectives to tasks (Article 5(3))

To deliver these objectives, Article 5(3) assigns four tasks: helping organisations accelerate digital transformation by connecting them with European providers of cloud and AI technologies (a); ensuring or providing access to upskilling and reskilling in close collaboration with the AI Skills Academy (b); facilitating the transfer of expertise across regions (c); and supporting the scaling-up of spin-offs and start-ups by facilitating access to clients (d).

Coverage and procedure

Article 5(1) requires each Member State to establish Centres for AI (recital context adds the aim of ensuring appropriate territorial coverage). Article 5(4) lets the Commission adopt implementing acts on the establishment procedure, participant-organisation profiles and selection criteria, under the examination procedure in Article 46(2). At the level of CADA's objectives, the Centres deliver operational objective 8 (regional and local adoption) named in Article 3(2)(h) and detailed in Article 4(8).

How the objectives connect to the wider framework

The three objectives are not free-standing. Each is reinforced elsewhere in CADA:

  • Integration and scaling (objective (a)) aligns with operational objective 5 (industrial AI) and operational objective 7 (public-sector AI), which the Initiatives pursue at Union level; the Centres bring that work down to the regional and sectoral level.
  • Broad adoption (objective (b)) is the regional and local face of operational objective 8 (Article 4(8)). The recitals describe the Centres as "regional and local accelerators for the uptake and deployment of AI," reinforcing the competitiveness and resilience of the Union's AI industrial base.
  • Leveraging infrastructure (objective (c)) connects to the network under Article 5(6), which can supply specialised services and compute "across regions where the required skills or compute capacity are not available locally," and to cooperation with other Union networks under Article 5(7), including initiatives in semiconductors and data.

The Centres also feed national policy: Article 7(2)(b) requires national cloud and AI strategies to support the Centres "as entry points to the European AI innovation ecosystem," so the objectives in Article 5(2) are meant to be picked up and resourced through each Member State's strategy.

What this means for you

For public-sector procurement officers and local government leaders, the Centres' objectives translate into concrete support. The objectives bind the Member States and Centres, not you directly, but they shape services you can use.

  • Localised expertise. Objective (a) means your regional Centre is mandated to help integrate and scale AI in public-sector settings — useful for smaller authorities with thin technical teams.
  • The "AI first" principle in practice. Use the Centres to embed considered AI assessment when redesigning services or drafting specifications — favouring scalable solutions, while weighing risks.
  • Scaling pilots. Objective (a) supports moving proven pilots to production, including connecting you with European cloud providers and the infrastructure to handle larger workloads.
  • Skills. Article 5(3)(b) routes reskilling through the Centres and the AI Skills Academy.
  • Supporting local innovation. Objective (b) and Article 5(3)(d) point to SMEs and start-ups you can engage through innovation procurement.

Common misconceptions

"The Centres are new, standalone entities." Article 5(1) states they build on the existing EDIHs — a refocusing of existing hubs, not necessarily new organisations.

"'AI first' means AI must be used in every public service." The principle is about considering AI's opportunities and risks when designing processes and services — not mandating AI everywhere.

"The Centres provide free models and unlimited compute." Objective (c) lets them leverage infrastructure to support model development and fine-tuning, but this is facilitation and access support — not an unconditional, free resource.

"Only large cities benefit." Objective (b) explicitly targets regional and local adoption; the design aims to avoid a digital divide between hubs and peripheral regions.

Related

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