Shengxing YANG, a Ph.D. student of the RITM laboratory at Université Paris-Saclay, supervised by Professor Ahmed Bounfour, will defend his Ph.D. thesis entitled “Architecting AI Competences Toward 2035: A Mixed-Methods Study on AI Issue Space Mapping and the Articulation of an Organizational Competence Taxonomy,” on February 20, 2026. The PhD defense will be held in the Gaudemet Room (Building D of the Jean Monnet Faculty).

Composition of the jury:

  • Thomas, J. Housel, Full Professor, Naval Postgraduate School (Information Sciences Department) – Reviewer
  • Didier Lebert, Professor, ENSTA – Reviewer
  • Hind Benbya, Full Professor, Western Sydney University – Examiner
  • Amélie Clauzel, Professor, Université Paris-Saclay (IUT de Sceaux) – Examiner
  • Valérie Fernandez, Professor, Institut Polytechnique de Paris – Examiner

PhD Thesis Abstract: This dissertation investigates how organizations can prepare for the coming wave of artificial intelligence (AI) transformation by 2035, with a particular focus on organizational competences. It adopts a multi-phase mixed-methods design combining a systematic literature review with Gioia-inspired coding, a semi-Delphi study, an international survey of intellectual property regimes, and complementary econometric modelling. A central outcome is the AI competences 2035 taxonomy as the blueprint and general heuristic for competence configuration, which identifies six competence categories: sector-specific domain expertise, technological (or material) competences, strategic and organizational competences, cognitive competences, interactional competences, and ethical and societal competences. Together with the IC-dVAL framework, the findings reveal that contextual AI competence architectures are sensitive to their environments and evolve dynamically over time (e.g., by decades, the AI project lifecycle lens, and further introduction of “special” AI competences). The research further uncovers two pressing challenges. First, the emergence of agentic AI intensifies the “Discretion Crisis” and raises concerns about workforce sustainability (as automation reallocates decision-making authority and compresses learning opportunities for human employees). Second, digital platforms increasingly act as essential competence framers, shaping the “grammar” of the implementation of AI. This platform dependency offers efficiency gains but also deepens risks of strategic lock-in for organizations. Theoretically, this dissertation extends the competence-based view into the domain of AI transformation by introducing the Issue Space, the AI competences 2035 taxonomy, contextual AI competence architectures, and the enriched IC-dVAL approach. Practically, it provides guidance for various stakeholders to articulate AI competences and identify risks. Additionally, this dissertation identifies the emerging phenomenon of “AI mediation” and proposes directions for future research. While limitations in methodology, data, theoretical framing, and the uncertainty of AI advancement, this work offers an early comprehensive framework for cultivating the competences needed to mitigate AI’s uncertainties and prepare for the AI transformation.