Exploring Teachers’ Lived Experiences in Assessing Authentic Student Learning in AI-influenced Classrooms: A Phenomenological Study

Abstract

Background and Aim: The increasing integration of artificial intelligence (AI) in education has transformed students’ approaches to academic tasks and challenged traditional assessment practices. While AI tools can support learning and productivity, they also raise concerns regarding the authenticity of student outputs and the validity of assessment outcomes. Grounded in Constructivist Learning Theory and authentic assessment perspectives, this study aimed to explore the lived experiences of senior high school science teachers in assessing authentic student learning in AI-influenced classrooms.

Materials and Methods: This study employed a qualitative phenomenological research design to examine teachers’ experiences in evaluating student outputs that may have been assisted by AI tools. Ten senior high school science teachers from selected public secondary schools in Misamis Occidental and Lanao del Norte, Philippines, participated in semi-structured interviews. Data were analyzed using Braun and Clarke’s six-step thematic analysis framework to identify recurring patterns, meanings, and themes related to assessment practices in AI-rich learning environments.

Results: The findings revealed five major themes: authenticity concerns, challenges in determining genuine understanding, assessment adaptation, professional judgment, and the redefinition of authentic learning. Teachers reported difficulty verifying whether polished and technically sound outputs genuinely reflected students’ understanding. Many participants observed inconsistencies between students’ written outputs and their oral explanations or classroom performance. In response, teachers adapted their assessment practices by incorporating oral questioning, in-class activities, performance-based tasks, and explanation-focused evaluations to make student thinking more visible. The findings also emphasized the importance of teachers’ professional judgment and ethical responsibility in maintaining fairness, academic integrity, and meaningful assessment practices in AI-influenced classrooms.

Conclusion: The study concludes that authentic learning in AI-rich educational environments can no longer be measured solely through written outputs. Instead, authentic learning is better demonstrated through students’ ability to explain, apply, and reflect on their understanding in meaningful contexts. The findings suggest the need for schools and educators to redesign assessment approaches, strengthen authentic and process-based evaluation, and establish clear guidelines for the responsible use of AI in education to ensure that technology supports rather than replaces genuine student learning.