The emergence of AI came with several positives and negatives. Its fair and ethical use has transformed education in several ways. However, there is a literature gap on how secondary school teachers in Saudi Arabia see the effect of the Integration of Generative AI (IGAI) on students’ performance in Physics (SPiP), with the mediating and moderating roles of class engagement and personalised learning (PL). The Constructivist Learning Theory supported it. The research design of the current study was a descriptive survey. A sample of 380 secondary school teachers from the Riyadh province was selected using multi-stage and multi-method sampling techniques. The data were collected online through Google Forms using a self-developed questionnaire. The analysis revealed that IGAI significantly contributes to and predicts SPiP. Class engagement was found to be a positive and significant mediator (80% of the total effect). Personalised learning was found to be a significant contributor to SPiP at the individual level. However, the interaction effect (IGAI*PL) was negative but significant for the relationship between IGAI and SPiP. By extending the implications of Constructivist Learning Theory, we found that IGAI is the need of the hour. Thus, teachers and students are advised to use AI fairly and ethically. The current study had several practical, policy, and research implications.