Exploring the Potential of Generative AI in Initial Teacher Training: A Motivational Analysis

Artificial Intelligence is reshaping teacher training by enhancing pedagogical practices through automation, personalized support, and intelligent content generation. As AI technologies integration is advancing globally, its adoption into in Moroccan teacher training remains constrained due to institutional resistance, insufficient training and lack of awareness. These challenges hinder future teachers’ engagement with GenAI technologies. This study examines the motivational dimensions influencing GenAI adoption among Moroccan future teachers, specifically ChatGPT, DeepSeek, and Grok, as intelligent supports for pedagogical task preparation during their initial training, using Keller’s Attention, Relevance, Confidence, and Satisfaction (ARCS) model and the Academic Motivation Scale (AMS). A quasi-experimental, quantitative approach was employed. Data were collected through a structured questionnaire based on the Attention, Relevance, Confidence, and Satisfaction-Instructional Materials Motivation Survey (ARCS-IMMS) and AMS components, administered to 146 future teachers enrolled in three distinct training specializations within ENS teacher training institution. Purposive and convenience sampling ensured disciplinary representation. Statistical analysis revealed that gender did not significantly affect motivation levels, as evidenced by an independent samples t-test (p = 0.403), with males reporting a mean score of 3.35 and females 3.41. The effect size, Cohen’s d = −0.156, indicated a small and practically negligible difference. Whereas, training specialization significantly influenced motivation (Fisher’s exact test, p = 0.046), with future teachers in literary disciplines reporting higher motivation (M = 3.49, SD = 0.385), likely due to the alignment between GenAI’s capabilities and language-related pedagogical tasks. Multiple regression analysis confirmed that components of both ARCS and AMS significantly predicted motivation levels (p < 0.001 for all variables). The model demonstrated high explanatory power, with a multiple correlation coefficient R = 0.987, indicating a very strong positive relationship between the motivational components and the overall motivation score. These findings highlighting the value of designing motivationally rich, cognitively engaging, and professionally relevant teacher training programs to support the effective pedagogical integration of GenAI tools. This study contributes to the growing body of literature on AI in education by addressing a gap in Moroccan teacher training. Further investigations are required to systematically evaluate its long-term impact across diverse educational settings.