Overview

• The study compared 145 essays written by university students with 145 essays generated by ChatGPT on the same topics.
• ChatGPT essays were found to be grammatically sound and coherent but lacked personal engagement strategies like rhetorical questions and direct reader appeals.
• Human-written essays were more interactive and persuasive, often demonstrating a clear stance and richer engagement techniques.
• Researchers highlight the need to treat AI as a teaching tool rather than a shortcut, emphasizing its inability to replicate critical thinking and personal voice.
• The findings underline ongoing concerns about academic integrity and the challenges of detecting AI-generated content in educational settings.