Artificial intelligence (AI) is fundamentally reshaping the landscape of university assessments, presenting what experts are calling a “wicked problem.” As generative AI tools like ChatGPT become more sophisticated, students can produce essays and answers in mere seconds, leaving educational institutions grappling with how to maintain academic integrity while adapting to this rapid technological evolution. Our recent research highlights that the challenges posed by AI in assessments are not just about finding a one-size-fits-all solution.

OpenAI’s new study mode for ChatGPT aims to enhance student engagement by employing the Socratic method, yet it raises concerns about the potential for students to bypass genuine learning. By encouraging users to think critically and respond to open-ended questions, this feature attempts to transform ChatGPT from a mere answer machine into a more interactive tutor. However, the challenge remains: with the ease of switching back to direct answers, will students truly engage with the material, or will they succumb to the temptation of quick fixes?