The Lived Experiences of English Teachers in Assessing Essays with Suspected or Disclosed AI Involvement: Toward an AI-Informed Assessment Framework

This descriptive phenomenological study examined how eight purposively selected English teachers at a public secondary school in Manila assessed essays with suspected or disclosed generative AI involvement. Semi-structured interview data were analyzed using Colaizzi’s method. Five themes emerged: recognizing AI through discrepancies in students’ writing, employing verification strategies before assessment decisions, negotiating assessment fairness and validity in AI-mediated writing, constructing authenticity through student ownership and personal expression, and adapting to the reality of AI while seeking support and guidance. Teachers compared submitted essays with students’ demonstrated abilities and sought clarification through student dialogue, comparison with earlier work, and, when needed, supervised writing. Teachers questioned whether grades represented student learning and sought clearer criteria. Authenticity involved independent performance, lived experience, personal voice, and student contribution. The experience was characterized by continuing negotiation of uncertainty while protecting authenticity, fairness, assessment validity, and academic integrity. Interpreted through Messick’s Unified Theory of Validity, the findings informed an AI-Informed Assessment Framework for English Writing Assessment that connects recognition, verification, authenticity evaluation, assessment decisionmaking, and professional adaptation.