This study examines how teacher–student relationships and school engagement change across online and in-school learning, based on the experiences of 105 cross-regional secondary vocational students in China. Using questionnaire surveys, the study explores students’ perceptions and learning needs in both settings. The findings confirm that teachers play a consistently positive role in promoting student engagement across both online and in-school learning modalities. While affective engagement was higher during online learning, driven by stronger teacher responsiveness and improved student–teacher relationships, students reported increased pride in their schools after returning home, reflecting a renewed appreciation. In-school learning was associated with higher behavioral engagement and greater motivation, despite tensions over intensified academic tasks. Online learning facilitated cognitive engagement through easier vocabulary searches; nevertheless, poor home environments reduced motivation. Enhancing engagement may require offering students autonomy, valuing their input, and clarifying the relevance of the learning content.

The development and application of artificial intelligence (AI) has accelerated the process of education and teaching transformation. Although prior studies have examined the forms of integrating AI into education, insights into the effective factors impacting pre-service teachers’ AI-assisted instruction intention (AI-AII) are rather limited. Considering this gap, this study constructed a structural model among AI-AII, AI pedagogical content knowledge (AI-PCK), AI technological knowledge (AI-TK), performance expectancy (PE), effort expectancy (EE), social influence (SI) and facilitating conditions (FC). This model explains how teachers can use pedagogical and content knowledge to improve instructional practices, and how technology can support the implementation of the pedagogical and content knowledge. Data were collected from 1,391 pre-service teachers in China. Results of the modelling effort indicate that the pre-service teachers’ AI-PCK, EE, PE, SI and FC positively predict their AI-AII. However, pre-service teachers’ AI-TK did not exert effects on their AI-AII. These insights are important for educators and policymakers to consider in designing teacher education and professional development related to foster pre-service teachers’ behavioural intention to use AI in teaching.