Need satisfaction and need frustration in generative artificial intelligence use among higher education students: A scoping review

The rapid adoption of generative artificial intelligence (GenAI) among higher education students has intensified the need to understand its motivational and psychological implications. However, the literature remains theoretically fragmented, making it difficult to develop a coherent understanding of how GenAI relates to students’ basic psychological needs. This scoping review systematically mapped the evidence on need satisfaction and need frustration associated with GenAI use among higher education students using Self-Determination Theory (SDT) as an analytical lens. The review followed the PRISMA Extension for Scoping Reviews (PRISMA-ScR). A systematic search across three electronic databases identified 42 studies after screening, involving more than 14,000 higher education students from 21 countries. The synthesized evidence showed that autonomy and competence were more frequently associated with need satisfaction, particularly when GenAI was integrated into pedagogically structured learning activities, whereas relatedness remained the least understood and least consistently supported basic psychological need. Need frustration—most often inferred from proxy indicators such as over-reliance, cognitive offloading, and diminished self-regulation rather than measured directly—was more commonly identified in studies examining intensive or minimally guided GenAI use. Instructional design consistently emerged as a prominent boundary condition associated with the balance between need satisfaction and need frustration. The evidence base remains limited by the predominance of cross-sectional studies and the concentration of research among undergraduate students in East Asia and the Middle East. Overall, the synthesized evidence suggests that GenAI does not inherently satisfy or frustrate students’ basic psychological needs. Instead, its motivational consequences appear to depend on the interaction among AI tool design, instructional scaffolding, and individual learner characteristics.