AI-generated vs. student-crafted assignments and implications for evaluating student work in nursing: an exploratory reflection.

Abstract

Objectives: Chat Generative Pre-Trained Transformer (ChatGPT) is an artificial intelligence-powered language model that can generate a unique outputs, in reponse to a user's textual request. This has raised concerns related to academic integrity in nursing education as students may use the platform to generate original assignment content. Subsequently, the objective of this quality improvement project were to explore and identify effective strategies that educators can use to discern AI-generated papers from student-written submissions.

Methods: Four nursing students were requested to submit two versions a Letter to the Editor assignement; one assignment that was written by the student; the other, exclusively generated by ChatGPT-3.5.

Results: AI-generated assignments were typically grammatically well-written, but some of the scholarly references used were outdated, incorrectly cited, or at times completely fabricated,. Additionally, the AI-generated assignments lacked detail and depth.

Conclusions: Nursing educators should possess an understanding of the capabilities of ChatGPT-like technologies to further enhance nursing students' knowledge development and to ensure academic integrity is upheld.