The impact of artificial intelligence on education: Insights from the Republic of Serbia

Abstract

Artificial intelligence (AI) is becoming increasingly relevant to educational practice, raising questions about teachers' willingness and capacity to integrate AI tools into their professional work. This study examined factors associated with secondary school teachers' behavioral intention to use AI in the Republic of Serbia within an extended Technology Acceptance Model (TAM). The study focused on self-efficacy, AI-related anxiety, perceived usefulness, and perceived ease of use and additionally examined the associations of behavioral intention with perceived pedagogical impact and perceived changes in teachers' professional roles. Data were collected from 120 secondary school teachers using an online questionnaire adapted from the Teachers' Acceptance of Artificial Intelligence in Education (TAAI) instrument and analyzed using descriptive statistics, Pearson correlation analysis, and linear regression. Self-efficacy and perceived ease of use were significant positive predictors of behavioral intention to use AI, whereas AI-related anxiety and perceived usefulness did not make significant unique contributions in the corresponding regression models. Behavioral intention was positively associated with perceived changes in teachers' professional roles and perceived pedagogical impact, although both associations were small. The findings suggest that teachers' intention to use AI is more closely related to their confidence in using AI and its perceived ease of use than to their general perception of its usefulness. In the Serbian secondary education context, the results point to the importance of practical AI competencies and institutional support while also indicating that favorable perceptions and behavioral intention should not be equated with actual AI adoption. The study provides empirical evidence on teachers' acceptance of AI in a national educational context in which such research remains limited.