Integrating Artificial Intelligence into Institutional Quality-Assurance Audits: A Parallel-Panel Study in Ontario Colleges

抄録

Recent advances in artificial intelligence （AI） offer quality-assurance （QA） agencies new opportunities to streamline labour-intensive institutional audits. This study reports on the first systematic comparison between a traditional human audit panel and an AI Audit-Panel Assistant （APA） that reviewed the same evidence package for an Ontario college under the College Quality Assurance Audit Process（ CQAAP）. Operating in strict isolation, each “panel” completed the full seven-stage audit cycle prescribed by the Ontario College Quality Assurance Service （OCQAS）. Convergence was observed between panels with strong alignment in evaluations across 30 requirements. We discuss the efficiencies, objectivity, and ethical challenges of AI-supported audits and outline a research agenda for responsible integration of AI in higher-education QA.