The rapid use of artificial intelligence (AI) in teacher education raises important concerns about whether pre-service teachers will continue using AI despite emerging risks such as perceived AI brain-rot exposure. Therefore, this study examines how UTAUT-related stimuli, institutional support, and perceived AI brain-rot exposure influence the intention to continue using AI through trust in AI. This study employed a quantitative cross-sectional survey design involving 247 pre-service teachers enrolled in teacher education programmes in Indonesia, all of whom had prior experience using AI for academic or teaching-related purposes. Data were analyzed using Partial Least Squares Structural Equation Modeling. The results showed that performance expectancy and social influence significantly increased trust in AI, whereas effort expectancy and institutional support did not significantly influence trust. Perceived AI brain-rot exposure also significantly influenced trust in AI, but the relationship was positive, suggesting that awareness of AI-related cognitive risks may coexist with selective or calibrated trust. Trust in AI strongly influenced continuance intention and mediated the effects of performance expectancy, social influence, and perceived AI brain-rot exposure on the continuance intention. The model explained 72.1% of the variance in trust in AI, and 62.6% of the variance in continuance intention. This study contributes to the literature by extending the UTAUT and S–O–R with a stressor perspective and by introducing perceived AI brain-rot exposure as an emerging construct in AI-in-education adoption research. These findings suggest that teacher education programmes should prioritize demonstrating AI's concrete pedagogical benefits and fostering reflective AI literacy to build trust, rather than relying solely on institutional policy or ease-of-use considerations.