Distinguishing Fact from Fiction: Student Traits, Attitudes, and AI Hallucination Detection in Business School Assessment

AI hallucination detection is a critical skill gap: Only 20% of second-year business students successfully identified factually incorrect AI-generated responses in a high-stakes assessment, highlighting that most students cannot reliably distinguish AI-generated falsehoods from facts—a concerning finding as AI becomes more prevalent in workplaces.

Critical thinking matters more than memorization: Strong predictors of hallucination detection included interpretive skills, writing proficiency, and AI scepticism, while "rote knowledge application proved less effective," suggesting that education systems should prioritize analytical reasoning over memorization to prepare students for an AI-integrated world.

Structured feedback can level the playing field: The study found that "structured feedback mitigates initial disparities in detection ability," indicating that educators can help students improve their AI literacy through targeted interventions rather than accepting detection ability as fixed.

AI scepticism is a valuable trait: AI scepticism emerged as a key predictor of hallucination detection ability, suggesting that healthy questioning of AI outputs—rather than uncritical acceptance—is an important component of responsible AI engagement.

Gender differences warrant attention: The research observed gender differences in detection ability, though the abstract does not specify the direction or magnitude, suggesting educators should examine whether certain groups need additional support in developing AI literacy skills.