A diagnostic human agency framework for AI driven education from a PRISMA guided systematic review

Artificial intelligence is rapidly transforming education and workforce development, yet the dominant discourse prioritises technical efficiency over the qualitative dimensions of human agency—autonomy, dignity, and the capacity to shape one’s learning and career trajectories. This article proposes the Human Agency Framework (HAF), a conceptual framework intended to support the evaluation of AI-driven educational initiatives through the lens of agency. Grounded in a systematic literature review (35 sources) and a novel “diagnostic design synthesis” methodology that integrates academic evidence with practitioner insights, the HAF operationalises agency across four interconnected dimensions: purpose (what valued capabilities learners gain), control (who directs learning pathways and data), collaboration (how human‑AI teams function), and conditions (the structural and policy enablers). The framework offers a potential lens for identifying tensions between efficiency claims and genuine empowerment, providing tentative guidance for educators, industry leaders, and policymakers. Two illustrative case studies—a UNESCO‑Huawei capacity‑building project and a global EdTech career guidance platform—illustrate how the HAF may be applied to examine potential strengths and weaknesses. By shifting the focus from what AI can do to what learners and workers can do and be with AI, The framework is conceptually aligned with the objectives of SDG 4 (Quality Education) and SDG 8 (Decent Work and Economic Growth). We conclude with a research agenda for empirical validation and cultural adaptation.