Faster, but Not Deeper? Generative AI and the Sustainability of Learning in Turkish Upper-Secondary Education: A Phenomenographic Study

Generative artificial intelligence (GenAI) has become an everyday study companion for secondary students, who use it intensively, largely outside any formal pedagogical design, within examination-driven systems that reward efficiency. Whether such use sustains or erodes the durable learning capacities that quality education (SDG 4) is meant to build—and how students themselves understand this trade-off—remains little explored. This phenomenographic study examines the qualitatively different ways in which twenty Grade 11 students in Türkiye conceive of GenAI within their self-directed studying, through semi-structured interviews on purposes of use, perceived learning value, ethics, and the tool’s imagined absence. The outcome space comprises one universal base conception—GenAI as an accelerated information channel—elaborated along two dimensions: the locus of agency in the instructional relationship, and the temporal horizon within which students evaluate the tool’s effects on themselves. Skilled governance of the tool and awareness of its long-term costs were not positively associated. Ethical reasoning was tiered and anchored in personal effort; no participant cited fear of being caught. Most imagined its absence as a mere slowdown; a few described dependence, motivation, or feasibility. Implications are discussed for AI-literacy curricula, process-based assessment, and equitable AI policy.