Students’ perceptions of multiliteracies development using AI-assisted portfolio assessment

In higher education, Generative Artificial Intelligence (GenAI) has opened new possibilities for developing student multiliteracies and set new requirements for multiliteracies pedagogies. While research has investigated GenAI's affordances and limitations of GenAI in language teaching, empirical studies on integrating GenAI in formative assessment practices to empower students' multimodal competence development remain lacking. This study examines how students develop multiliteracies throughout e-portfolio compilation in an EFL, GenAI-supported portfolio assessment classroom and uncovers students' perceptions of using GenAI in formative assessment. Employing a qualitative design over 16 weeks, 50 freshmen students were divided into groups to collaboratively create WordPress-based portfolios featuring multimodal writing about touring routes recommendations. Students reflected and revised their multimodal artefacts based on self, peer, teacher and/or AI-generated feedback. Students' reflective journals, focus group interviews and narrative inquiries were collected and analysed thematically. Findings showstudents develop multiliteracies through three stages, moderate transformation from passive to proactive learning, growing confidence and multimodal awareness, enhanced metacognitive development and professional growth. Students' perceived affordances and challenges are discussed. This study contributes to portfolio assessment scholarship by introducing GenAI as a key component in assisting students' feedback and assessment practices. Pedagogical implications are proposed for optimizing GenAI use to benefitstudents' learning.