Summary Under the proposed Cloud and AI Development Act (CADA), public-sector contracting authorities must include non-price award criteria evaluating a tenderer's contribution to the European cloud and AI ecosystem. As proposed in Article 32, these "Union added value" criteria are mandatory for innovative cloud and AI procurements but must remain ancillary and not decisive. To draft them correctly, you must assess four specific dimensions: strengthening the digital supply chain, integrating EU technologies, innovation for security of supply, and the use of EU-designed hardware (where feasible). Crucially, the weighting must be modest to respect the "ancillary" constraint, ensuring that technical quality and price remain the primary drivers of the award decision.

Detail

The proposed Cloud and AI Development Act (CADA), COM(2026) 502 final, fundamentally shifts the logic of public procurement in the cloud and AI sector. While traditional procurement focuses on the "most economically advantageous tender" (MEAT) based primarily on price and technical performance, CADA introduces a strategic layer designed to reduce dependencies and foster technological sovereignty. Article 32 establishes the legal framework for "Union added value," transforming procurement from a passive purchasing activity into an active instrument of industrial policy.

This obligation applies specifically to public procurement procedures for innovative cloud computing services and AI systems. It is not a voluntary option; contracting authorities "shall include" these criteria as part of the quality evaluation. However, the proposal is meticulously crafted to prevent these criteria from becoming disguised protectionism or a violation of the fundamental principles of the internal market. The criteria must be objective, transparent, and subordinate to the core requirements of the contract.

The Four Pillars of Assessment: Article 32(3)

When drafting your tender documents, you cannot rely on vague concepts of "European preference." Article 32(3) explicitly defines four distinct dimensions that contracting authorities must enable themselves to evaluate. Your award criteria must map directly to these pillars to ensure legal compliance and effective assessment.

  1. Strengthening the Digital Supply Chain (Article 32(3)(a)) The first dimension requires evaluating the extent to which the tenderer contributes to strengthening the digital technology supply chain in the Union. This is not limited to the final assembly of a product. The text explicitly includes "the use of software or hardware designed or manufactured in the Union."

    • Drafting implication: You should ask tenderers to disclose the origin of their core software components and hardware. Evidence could include declarations of design locations, manufacturing sites, or supply chain mapping showing a reduction in reliance on single points of failure in third countries.

  2. Integration of EU Technologies (Article 32(3)(b)) This dimension assesses whether the tenderer has integrated technologies developed in the Union. The scope is broad, covering "research and development results stemming from Union funded research and development programmes." It also extends to the use of "tools, such as standards, specifications, software, models or other technology developed in the Union."

    • Drafting implication: Your criteria should reward the use of EU-funded R&D outputs (e.g., models developed under Horizon Europe or Digital Europe Programme) and adherence to EU-developed standards. This encourages the uptake of European innovation and ensures public spending circulates within the EU R&D ecosystem.

  3. Innovation for Security of Supply (Article 32(3)(c)) This is a forward-looking criterion. You must evaluate whether "the innovation required to deliver the service contributes to strengthening the security of supply and the development of a European cloud and AI ecosystem."

    • Drafting implication: This moves beyond static "where is it made" questions to dynamic "how does this help the EU" questions. Tenderers might be scored on their plans to localize critical capabilities, their resilience against supply chain shocks, or their contribution to the development of sovereign alternatives.

  4. Hardware Origin and Feasibility (Article 32(3)(d)) This is the most technically specific and nuanced pillar. You must evaluate whether the service is delivered, "to the greatest extent feasible with regard to market availability and technical requirements," through critical computing, storage, and networking hardware components designed and/or manufactured in the Union.

    • Crucial Caveat: The proposal acknowledges market realities. If EU-designed hardware is not feasible, the criteria allow for the evaluation of hardware from a third country, provided it "contributes to strengthening the security of supply and the development of a European cloud and AI ecosystem."
    • Drafting implication: You must include a "feasibility" clause in your criteria. Do not set an absolute requirement for 100% EU hardware if the market cannot supply it. Instead, ask tenderers to demonstrate the maximum feasible use of EU hardware and, where they must use third-country components, explain how those components still support the EU ecosystem (e.g., through local assembly, R&D investment, or security features).

The "Ancillary" Constraint: Article 32(2)

The most critical legal constraint in drafting these criteria is found in Article 32(2). The proposal sets out four strict conditions that your non-price award criteria must satisfy. Failure to adhere to these conditions could render the procurement procedure legally vulnerable.

  • Linked to the Subject Matter (Article 32(2)(a)): The criteria must be "linked to the subject matter of the contract." You cannot use Union added value to reward a tenderer for unrelated activities, such as general corporate social responsibility or unrelated charitable donations. The contribution to the EU ecosystem must be inherent in the delivery of the specific cloud or AI service being procured.
  • No Unrestricted Freedom of Choice (Article 32(2)(b)): The criteria must "not confer unrestricted freedom of choice on the contracting authority." This means the scoring methodology must be objective, measurable, and transparent. You cannot reserve the right to decide subjectively which tenderer has the "best" EU value after the bids are opened. The rules for scoring must be predefined and applied consistently.
  • Expressly Set Out in Procurement Documents (Article 32(2)(c)): The criteria must be "expressly set out in the procurement documents or in the contract notice." Surprise criteria introduced during the evaluation phase are strictly prohibited. Tenderers must know exactly how they will be scored on Union added value before they submit their bids, allowing them to tailor their offers accordingly.
  • Ancillary and Not Decisive (Article 32(2)(d)): This is the most important safeguard. The non-price award criteria relating to Union added value must be "ancillary and not decisive in the award of the contract." They cannot override the core technical and financial performance requirements. If a tenderer offers a superior technical solution at a lower price but scores lower on Union added value, they should not automatically lose the contract solely due to the ancillary criteria. The primary focus must remain on the quality and cost-effectiveness of the service itself.

Weighting and Scoring Strategy

While Article 32 does not prescribe a specific numerical weighting, the requirement that the criteria be "ancillary and not decisive" implies that the weighting must be modest. In practice, this means the Union added value criteria should not dominate the overall score.

The explanatory memorandum accompanying the proposal suggests a practical benchmark: a maximum weighting of 15 out of 120 points (12.5%) allocated to European added value within the overall evaluation methodology. This ensures that the criteria remain proportionate and subordinate to the core contract award criteria.

  • Recommended Approach: Allocate a specific, modest percentage (e.g., 10–15%) of the total quality score to Union added value.
  • Scoring Logic: Ensure that the scoring rubric is granular. For example, award points for specific evidence (e.g., "5 points for providing a full SBOM showing >50% EU-origin components," "3 points for using EU-funded R&D models").
  • Avoiding Decisiveness: Structure your evaluation so that a high Union added value score cannot compensate for a failure to meet the minimum technical requirements or a significantly higher price. The "ancillary" nature means these criteria act as a differentiator among qualified, competitive tenders, not as a gatekeeper that excludes the most economically advantageous tender.

What this means for you

For public-sector procurement officers, drafting Union added value award criteria under CADA requires a shift from purely technical and financial evaluation to a more strategic assessment of supply chain resilience and technological sovereignty.

1. Integrate Early in the Process Do not treat Union added value as an afterthought. When defining the technical specifications for your cloud or AI procurement, consider how the four dimensions in Article 32(3) can be measured. For example, if you are procuring AI training services, ask tenderers to specify the origin of the hardware used for training and whether they utilize EU-developed AI models or frameworks.

2. Define Clear Scoring Metrics Because the criteria must be objective and not confer unrestricted freedom of choice, you need clear metrics. For instance, you might score higher on a tenderer that provides a detailed Software Bill of Materials (SBOM) showing a high percentage of EU-origin components, or one that demonstrates the use of EU-funded research outcomes. Avoid vague language like "contributes to the EU ecosystem" without defining what evidence will be accepted.

3. Balance with Technical Requirements Remember the "ancillary" constraint. Ensure that your overall scoring model prioritizes the technical performance and price of the service. Union added value should be a tie-breaker or a modest bonus, not the primary determinant. If a tenderer cannot meet the technical requirements, their Union added value score should not save them.

4. Prepare for Evidence Submission Tenderers will need to provide evidence to support their claims. Your procurement documents should specify what kind of evidence is acceptable. This might include certificates of origin, SBOMs, documentation of R&D funding sources, or declarations on hardware design and manufacturing locations. Make it clear that unsubstantiated claims will not be scored.

5. Monitor and Report CADA also requires Member States to monitor their use of procurement of innovation in cloud and AI (Article 33). Keep records of how you applied the Union added value criteria and the outcomes. This data will be useful for reporting on SME participation and the overall impact of the procurement on the European cloud ecosystem.

Common misconceptions

Misconception 1: Union added value is a quota for EU content. Reality: Article 32 does not impose a strict quota. It allows for the evaluation of contribution to the EU ecosystem. As seen in Article 32(3)(d), if EU hardware is not feasible, third-country hardware can still be considered if it strengthens security of supply. The focus is on the contribution to the ecosystem, not just a binary EU/non-EU label.

Misconception 2: I can use Union added value to exclude non-EU providers. Reality: The criteria must be non-discriminatory and linked to the subject matter. While the goal is to strengthen the EU ecosystem, the criteria must be applied fairly. A non-EU provider that integrates significant EU technologies, uses EU-designed hardware, or contributes to EU security of supply can still score well. The criteria are not a blanket ban on non-EU participation.

Misconception 3: Union added value can be the deciding factor. Reality: Article 32(2)(d) explicitly states that these criteria must be "ancillary and not decisive." If a tenderer offers a significantly better technical solution at a lower price, they should not lose the contract solely because a competitor scored higher on Union added value. The primary criteria remain technical quality and price.

Misconception 4: I can decide the scoring method after receiving bids. Reality: Article 32(2)(c) requires that the criteria be expressly set out in the procurement documents or contract notice. You must define how you will score Union added value before tenderers submit their bids. This ensures transparency and fairness, allowing tenderers to tailor their bids to meet your specific requirements.

Related

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