Enhancing AIGC credibility assessment in higher education: an extended framework integrating user trust, AI literacy, and behavioral adoption

This study extends the foundational work of Cui and Zhang (2025. Can student accurately identify artificial intelligence generated content? An exploration of AIGC credibility from user perspective in education. Education and Information Technologies, 30 (12), 16321–16346. https://doi.org/10.1007/s10639-025-13448-1) by examining AI-generated content (AIGC) credibility assessment among 286 higher education students across three Chinese universities. Employing an extended Technology Acceptance Model (TAM) framework integrated with AI literacy and critical thinking constructs, we investigated factors influencing students’ ability to distinguish AIGC from human-created content. Results revealed that participants achieved 68.4% accuracy in general AIGC recognition but only 57.9% for domain-specific content. Structural equation modelling demonstrated that perceived usefulness (β = .44) and ease of use (β = .31) significantly predicted attitudes towards AIGC (R2 = .72), while interpretability (β = .63) and ease of use (β = .29) explained 91% of variance in AI illusion. AI literacy mediated the relationship between perceived usefulness and attitudes (β indirect = .07, p = .008), while critical thinking mediated the perceived ease of use–AI illusion pathway (β indirect = −.05, p = .024). These findings inform the development of comprehensive training programs for appropriate AIGC credibility assessment and trust calibration in educational contexts.