Artificial Intelligence (AI) is becoming increasingly central to public sector transformation in Europe, yet its adoption poses significant challenges for capacity building, ethical deployment, and cross-role collaboration. The AI4Gov-X initiative addresses these challenges through an integrated ecosystem of training and digital tools, underpinned by a competency framework tailored to governance realities. This workshop is the first in a series of validation activities designed to test and refine that framework. Using interactive methods, participants will assess the clarity and relevance of functional profiles, validate and prioritise competencies, and identify gaps or emerging skills. The session will generate directly usable evidence-validated lists, ranked priorities, and proposals for refinement-that will inform both academic publications and the further development of AI4Gov-X tools. By engaging a diverse group of stakeholders, the workshop will enhance institutional capacity for the responsible and democratic adoption of AI in public administrations.
Artificial Intelligence (AI) is central to public sector transformation in Europe, yet its adoption raises
challenges for capacity-building, ethics, and cross-role collaboration [
The AI4Gov-X initiative addresses these through an ecosystem of training and digital tools, built on a
governance-oriented competency framework [
The 90-minute workshop has four phases. Each is designed to generate evidence for framework Opening (15 min). Welcome and objectives by the Scientific Director of AI4Gov-X, Gianluca Misuraca. Participants will briefly indicate their roles and governance levels to contextualise the findings. The competency framework and profiles are briefly presented to provide a shared basis for CEUR Workshop
ISSN1613-0073
Activity 1: Validating clustering (35 min). This activity will test the robustness of the framework’s clustering methodology. Participants will individually review the five functional profiles, assessing their clarity, completeness, and distinctiveness. Using the Slido platform to ensure anonymity, they will identify potential shortcomings—such as overlaps, complexity, or missing dimensions—and propose refinements or alternative groupings. The activity will conclude with a voting exercise to prioritise areas requiring urgent revision. The outcome will be direct evidence on the validity and usability of the clustering, complemented by suggestions for structural improvements.
Activity 2: Validating competencies (25 min). The second activity will assess the validity and clarity of the competencies linked to each functional profile. Working in groups, participants will review the competencies profile by profile, annotating items as relevant, unclear, or missing. They will then prioritise the most critical competences through a dot-voting exercise. This process will generate a ranked and validated list of competencies, highlighting those that require refinement, and identify gaps or emerging skills to be incorporated into the framework.
Final discussion (15 min). The closing plenary will synthesise insights from the previous activities. Participants will collectively identify urgent gaps, clusters requiring revision, and opportunities for new training approaches or institutional applications. The discussion will conclude with a prioritisation of the most critical competencies, resulting in consolidated recommendations and a roadmap for further development of the framework.
The session will yield: (i) a clear mapping of gaps and strengths within the framework; (ii) evidence on the validity and usability of the clustering methodology; and (iii) refinements to both the overall structure and the competencies identified. Broader stakeholder involvement will ensure the framework’s relevance, usability, and impact, reinforcing AI4Gov-X as a reference for capacity building in digital governance.
The authors utilised Grammarly for language refinement, including sentence structure and word choice, and afterwards, they carefully reviewed and edited the manuscript, taking full responsibility for the final version.