Foreign language media literacy as a protective factor against AI-generated disinformation and psychological stress in technical higher education in Ukraine

Abstract

The study addresses the twin problem of AI-generated disinformation in English-language technical sources and the psychological strain this creates for Ukrainian students studying under conditions of war. We model foreign language media literacy – FLML – as a protective factor that reduces epistemic error while moderating stress during time-bounded tasks. Drawing on inoculation theory, cognitive load theory, dual-process accounts, and appraisal-coping perspectives, we derive a mechanism that links compact verification micro-skills to improved judgement and calmer task appraisals. We designed a field-ready curriculum with three modules – detection, verification, and appraisal-coping – an embedded it across laboratory and project work in power engineering, electronics, and information systems. Implementation acknowledged wartime realities through a layered defence approach that included offline materials, a pause-and-resume protocol for air-raid alerts, and paper-digital verification logs to ensure continuity. Evaluation combined pretest-posttest-retention measures of verification accuracy, time-to-verification, and justification quality with brief ratings of cognitive load and perceived stress, and with process traces that recorded steps actually taken. Results show clear gains in verification accuracy and justification quality, increased use of lateral reading and source triangulation, and stable or slightly reduced time-to-verification under realistic constraints. Students reported a reliable shift threat to challenge appraisal, with lower perceived stress during timed verification, and the schema supported fast recovery after interruptions without loss of rigour. Effects were stronger for higher L2 proficiency and greater domain knowledge, highlighting the value of explicit scaffolds, bilingual glossaries for modality and evidence, and routine AI-use disclosure coupled with verification rather than prohibition. We conclude that FLML constitutes a practical, scalable shield for technical higher education in Ukraine – hardening epistemic defences while sustaining mental health – and we provide an implementation blueprint that invites replication and longitudinal tracking across programmes.