Beyond Behavioral Intention: A Context-Validated Framework for Meaningful Use of PictureThis AI in Ugandan Public Universities

Abstract

Purpose: This study validated and refined a context-specific framework explaining meaningful use behavior of PictureThis artificial intelligence (AI) for plant identification among undergraduate students in Ugandan public universities. Unlike an earlier analysis of predictors of behavioral intention from the same doctoral study, this article centres on the intention-to-use gap, actual academic engagement, mediation, moderation, and framework refinement. Materials and Methods: A positivist, quantitative, cross-sectional correlational survey was conducted with 477 biological and agricultural science students from Makerere, Busitema, Gulu, and Kabale universities. After at least 30 days of guided PictureThis use, participants completed an adapted UTAUT2 questionnaire. Use behavior was operationalized as user cognitive absorption and deep-structure usage. Measurement and structural models, mediation, and moderation were evaluated with partial least squares structural equation modelling in SmartPLS 4. Findings: The seven UTAUT2 predictors explained 62.7% of behavioral intention, but intention explained only 0.7% of meaningful use and did not significantly predict it (β = .081, p = .252). Hedonic motivation showed the clearest route to actual use, with a significant direct effect (β = .143, p = .008) and a significant indirect pathway through intention. Habit strongly predicted intention and retained an indirect pathway, whereas facilitating conditions were non-salient. Age moderated habit–intention and gender moderated price value–intention; most other demographic interactions were negligible. Implications to Theory, Practice and Policy: The refined framework qualifies UTAUT2 by separating intention formation from meaningful academic use. Universities should move beyond awareness and stated willingness by embedding repeated, enjoyable, guided, and assessable AI-supported plant-identification tasks, while maintaining botanical verification, equitable access, privacy, and academic-integrity safeguards. Keywords: Picturethis AI; Use Behavior; UTAUT2; Plant Identification; Higher Education