Exploring Current Practices and Challenges of Using Generative AI -ChatGPT-for Personalized Learning: A Case Study for High School Science Education

Generative AI learning has brought new possibilities of personalized inquiry-based learning. This study examines the uses, issues, and guidelines of the current implementation of ChatGPT in science lessons of high school students in the Institute of applied Technology (IAT) in the United Arab Emirates. The study was conducted against a constructivist framework and based on Straussian Grounded Theory, and it involved semi-structured interviews with 15 science teachers, generating 1074 coded data segments. The five essential themes which appeared were: AI literacy and competency, curriculum alignment, policy and stakeholder coordination, infrastructure equity, and ethical AI use. The systemic obstacles to the broader use of ChatGPT were the inequalities in policies, inadequate training, and access that accompanied the innovative use of ChatGPT as teachers differentiated and assessed continuously. To fix this, the study suggests the use of the PRISM-AI Framework (Policy, Resources, Integration, Support, Monitoring), which is a multi-level framework and offers application strategies to students, teachers, schools, and ministries. The findings stress the importance of the standardized, ethical, and inclusive practices of AI according to the UAE National AI Strategy 2031. The present study contributes a contexted regionally, and practically applicable framework to guide the addition of AI in science education at the K-12 level and fill the gap between the innovation promise and classroom delivery.