AI-Driven Assessment and Feedback in Work-Integrated Learning: A Systematic Review of Authenticity, Ethics, and Professional Competence

Abstract

Artificial intelligence (AI) is rapidly transforming the methods by which higher education institutions assess learning and provide feedback; however, its implications for work-integrated learning (WIL), wherein assessment must accurately reflect authentic professional performance, remain under-theorised. This systematic review synthesises evidence on AI-based assessment, automated feedback, learning analytics, and competency evaluation as they pertain to authenticity, ethics, and professional competence in WIL and related higher education contexts. Following the PRISMA 2020 guidelines, five databases (Scopus, Web of Science, ERIC, EBSCOhost, and the ACM Digital Library) and supplementary citation searching yielded 1,175 records. After the removal of duplicates and a two-stage screening process, 20 studies published between 2017 and 2025 were included and synthesised narratively in relation to four review questions. Findings indicate that AI tools can enhance the efficiency, scalability, and timeliness of feedback and support personalisation, particularly for the reflective and formative writing tasks that are central to WIL. However, the same tools raise persistent concerns: threats to assessment authenticity and academic integrity from generative AI, demonstrable algorithmic bias against linguistically and culturally diverse learners, a lack of transparency that undermines clarity, and the risk of over-automation that displaces the situated human judgement essential for professional competence. The review argues that AI should augment rather than replace evaluative judgement, and that authentic WIL assessment requires human-in-the-loop designs, validity-centred reform, and explicit attention to equity. Implications for assessment design, policy, and future research are discussed.