Teachers’ perceptions and challenges in the use of Artificial Intelligence (AI) in teaching science

This study examined the perceptions and challenges of public school science teachers regarding the integration of Artificial Intelligence (AI) in science education. Specifically, the research assessed teachers' perceived usefulness, perceived competence, professional development and training, and perceived challenges to AI adoption, while investigating the relationships among these variables, their predictive influence, and potential differences across demographic profiles. Adopting a quantitative descriptive-correlational design, data were collected from 113 science educators through a validated 40-item survey instrument. Statistical analyses included weighted means, Pearson r, multiple regression analysis, and differential tests (t-tests and ANOVA).Findings revealed that while teachers demonstrated a highly positive level of perceived usefulness (WM = 3.42, SD = 0.62), their perceived competence (WM = 3.14, SD = 0.62) and access to professional development and training (WM = 2.72, SD = 0.62) were only moderately developed. Additionally, respondents reported high levels of perceived challenges (WM = 3.02, SD = 0.60), specifically in the areas of data privacy, infrastructure limitations, and ethical concerns. Significant correlations were established between perceived competence, professional development and training, and perceived usefulness. Multiple regression analysis further indicated that competence and training significantly predicted perceived usefulness, accounting for 53% of its variance (R2 = 0.53). No significant differences were found across demographic variables. These results indicate that while teacher receptivity is high, effective AI integration in science teaching requires addressing technical gaps through structured professional development and training and mitigating systemic perceived challenges to ensure responsible and sustainable adoption.