Exploring Mathematics Teachers’ Experiences in Designing and Implementing AI-Generated Assessments

Abstract

This qualitative phenomenological study explored the lived experiences of mathematics teachers in designing and implementing AI-generated assessments in public and private high schools across Lanao del Sur, Philippines. Grounded in the Unified Theory of Acceptance and Use of Technology (UTAUT) and Pedagogical Content Knowledge (PCK), the research examined teachers’ perceptions of value, design processes, encountered challenges and opportunities, and strategies for ensuring alignment with instructional goals. Data were gathered through semi-structured interviews with six purposively selected teachers and analyzed using Moustakas’s modified Stevick-Colaizzi-Keen method. Findings revealed that teachers view AI primarily as a supportive tool rather than a replacement for professional expertise, valuing it most for reducing workload and generating diverse assessment materials. They follow a deliberate workflow involving clear prompt specification, rigorous verification, contextual localization, and iterative refinement. Key challenges include mathematical inaccuracies in AI outputs and unstable internet connectivity, while opportunities lie in enhanced efficiency and support for differentiated instruction. Teachers ensure quality by anchoring assessments in curriculum standards and adapting content to the cultural and learning needs of local students. The study concludes that effective AI integration follows a “human-in-the-loop” model, where technology serves as an aid while teachers retain authority over pedagogical quality. Recommendations are offered for teachers, school leaders, policymakers, developers, and future researchers to promote responsible and equitable AI use in mathematics education.