Experts warn students relying on AI for writing weakens critical thinking

Experts in the United States have raised concerns that students relying on AI to complete assignments or write entire essays may be weakening their critical thinking abilities. According to a report by The New York Times, the phenomenon of students using AI tools for academic work has become widespread, with some schools allowing limited use of such technologies. Cognitive psychologist Ronald T. Kellogg emphasized that writing is a technique for thinking, and the process of drafting a paper imposes significant mental effort, similar to physical labor. He argued that writing cannot be separated from the cognitive work it involves, and that delegating this task to AI means students lose the mental training that comes from the process. A study from MIT found that individuals who used AI to write essays showed lower brain activity compared to those who wrote without assistance, and struggled to recall their own words. Steve Graham, an expert in writing research at Arizona State University, warned that outsourcing any part of the writing process to AI comes at a cost. The New York Times highlighted that if writing is a form of thinking, then relying on AI to provide answers undermines students' ability to form their own judgments. In response, many schools in the U.S. are rethinking how to limit AI use, with some returning to paper-based writing and oral exams, while others have banned digital devices in classrooms. Princeton University and other Ivy League schools have also modified their honor codes to prevent students from taking exams without supervision. Some students have begun to recognize that using AI to write literary analysis essays means they are forgoing part of their educational development. One university student told The New York Times that without engaging deeply with difficult texts, they are not developing the necessary analytical skills. The article also references a separate report on how U.S. university professors are using hidden prompts in exams to detect AI cheating, with 32 out of 35 students caught using such methods.