Abstract:

This paper presents an empirically grounded framework for designing Artificial Intelligence (AI) learning across the curriculum through a capability-oriented approach. It builds on over a decade of the author’s research in learning design and rapid curriculum prototyping, and three years of work developing an AI Capability Framework informed by Martha Nussbaum’s Capability Approach. The paper addresses a crucial conceptual tension in education: the dichotomy between technical rationality and capability-oriented pedagogy. While technical rationality privileges efficiency, skills, and tool mastery, a capability perspective emphasises agency, identity, and human flourishing, positioning education as the expansion of learners’ freedoms to think, choose, and act responsibly in technologically mediated contexts.

Drawing on Nussbaum’s human capabilities—such as reason, imagination, affiliation, and control over one’s environment—the paper argues that a capability lens reorients AI education from the mastery of tools to the cultivation of ethical judgment, reflexivity, and social participation. AI thus becomes not only a subject of learning but also a context for exploring questions of human purpose, justice, and creativity. This reframing positions curriculum design as a process of enabling becoming, situating AI within broader debates on human‑centred, values‑based learning.

Methodologically, the study employs participatory design as both a research and pedagogical strategy. Educators and students acted as co-designers through iterative cycles of analysis, scenario creation, and prototype testing. Between 2023 and 2025, a series of design workshops were held in Greece and China across teacher education, business, and engineering. Scenario‑based activities elicited participants’ visions of AI learning, mapped desired learner capabilities, and explored institutional opportunities and constraints.

Two research questions guided the study:
(1) How can the Capability Approach provide a theoretical and pedagogical rationale for embedding AI across the curriculum?
(2) What institutional and pedagogical shifts are required to support AI learning that fosters agency, identity, and ethical insight alongside technical competence?

Findings indicate that integrating capability thinking enabled educators to reconceptualise AI literacy as a multidimensional construct integrating technical proficiency, ethical awareness, and human agency. Participatory design served as professional learning, promoting critical dialogue and enabling theoretical insights to be translated into teaching practice. However, structural tensions persisted between institutional accountability regimes—rooted in technical rationality—and capability‑based aims that resist standardisation. Cross‑cultural insights revealed the alignment between capability theory and Chinese educational traditions emphasising moral and collective development, yet also exposed challenges in balancing humanistic goals with performance‑driven metrics.

The study proposes an AI Capability Learning Design Model encompassing four dimensions: Technical Understanding, Ethical and Critical Reasoning, Creative and Collaborative Practice, and Reflexive Agency and Identity Formation. The model bridges the divide between technical rationality and capability development, arguing that both are necessary for AI education that prepares learners to shape technology in line with human values.

Keywords:

Capability-based learning, artificial intelligence, learning design.