Today, universities worldwide are actively embracing generative artificial intelligence (GenAI) tools represented by ChatGPT. A 2024 global survey indicated that a high percentage of 86% of university students frequently use AI in their daily studies.
The rapid proliferation of AI in the field of education has also sparked widespread debate. As reported by the UK’s Nature website, many believe that AI helps improve educational quality and assists students in coping with a rapidly changing world; however, many others worry whether over-reliance on AI might inhibit the development of students’ independent thinking and critical thinking skills.
AI Sweeps Through University Campuses
The proportion of young people using AI is remarkably high. This year, the AI company Anthropic analyzed 1 million anonymous conversations between university students and its generative AI tool “Claude,” finding that usage rates among students in Science, Technology, Engineering, and Mathematics majors were significantly higher than those in Business and Humanities majors.
What is the main purpose for students using these generative AI tools? A survey of over a thousand UK students by the UK Oxford University think tank “Higher Education Policy Institute” showed that most students use AI to assist with writing, optimizing, or summarizing texts.
Meanwhile, tech giants like OpenAI and Google are actively promoting their products to universities and students. Many universities are also proactively embracing this wave. A team from the University of Sydney in Australia developed a GenAI platform called Cogniti based on the ChatGPT interface. Since its launch in 2023, this platform has been integrated into the university’s digital teaching system, supporting teachers in creating customized AI assistants for different courses—it can serve as an intelligent tutor for science modules or transform brief grading comments into detailed academic feedback. Currently, over a thousand teachers university-wide are actively using the platform, and results are being shared with over a hundred universities globally.
In April last year, scientists at Tsinghua University constructed a three-layer AI teaching integration framework: the bottom layer incorporates AI models from several companies, including DeepSeek, Alibaba Cloud, OpenAI, and Google; the middle layer features a “knowledge engine” providing precise and cutting-edge information for various disciplines; the top layer integrates multiple student platforms, such as teaching systems equipped with AI teaching assistants and freshman guidance robots. It is reported that this system has been adopted by hundreds of universities in China.
Mixed Impact on Academic Performance
Do students who use AI tools truly acquire more knowledge? Preliminary research results have already raised warning signs. Research from Tsinghua University showed that students using AI learning companions performed better than non-users in initial post-lesson tests, but after two to three weeks, this advantage reversed, with AI users scoring lower than the control group.
According to the latest news, research from the MIT Media Lab, for the first time through brain science studies, revealed the impact of AI-assisted writing on brain activity. Their study, published at the end of June, showed that students who relied on ChatGPT to complete writing tasks had significantly reduced brain activity. This study resonated widely within the global education community—the researchers received 4,000 emails from primary, secondary, and university teachers worldwide, many expressing concerns about students’ over-reliance on AI.
Although this study had a limited sample size and did not consider long-term effects, it indeed revealed a common concern: students overly dependent on AI may be losing critical thinking skills—the core competency of deeply analyzing information and making rigorous judgments. Data from Anthropic’s analysis corroborates this worry: in their analysis of a million student-AI conversations, nearly half of the students directly asked for answers or ready-made content.
Professor Olivia Guest from Radboud University Nijmegen in the Netherlands also stated that in recent years she has clearly observed a decline in student abilities, with some students struggling to complete dissertations or write in-depth articles independently. In light of this, in June of this year, over 1,000 global experts published an open letter opposing the uncritical adoption of AI technology by universities. It is reassuring that some students have begun consciously reducing their AI usage, realizing that over-reliance could hinder intellectual growth.
Controversy Persists
However, Dr. Ravi B. from The Ohio State University believes these concerns might be exaggerated. He analogizes that just as calculators replaced basic arithmetic, delegating certain tasks to AI could free up the brain for higher-order thinking.
Some studies show that GenAI tools can help students learn. An experiment by Professor Greg K. ‘s team at Harvard University demonstrated that undergraduate students receiving tutoring from a customized AI tutor in physics significantly improved their learning efficiency. Among nearly 200 participating students, the median test score of the AI-tutored group jumped from a baseline of 2.75 (out of 5) to 4.5, far exceeding the 3.5 score of the traditional instruction group.
Professor K. believes the key lies in how it is used; AI can either replace thinking, making people lazy, or become a tool that promotes deep thinking. The core of education is whether the use of AI can genuinely stimulate students’ critical thinking and learning enthusiasm. Ms. Shafika E., Head of the Educational Technology and AI unit at UNESCO, reminds us that the application of various GenAI tools by teachers and students is continuously accelerating, and corresponding policies, teaching methods, and ethical standards urgently need to keep pace.
