By ZHANG QINGLI

and BAN XIAOYUE

In April this year, China’s Ministry of Education and four other departments jointly issued the “AI + Education” Action Plan, which explicitly states that by 2030, a pattern of deep integration between artificial intelligence (AI) and education will have basically taken shape as AI continues to reshape the landscape of university education.

The extent to which AI has entered university classrooms around the world has far exceeded what many people might imagine. Chen Xiangdong, a professor from the Faculty of Education at East China Normal University, cautioned against hastily labeling the use of AI as “dependence.” Human progress, he noted, has always involved expanding human capabilities through the use of new tools. The real questions are what people depend on, to what extent, and what irreplaceable abilities they retain after using such tools.

When AI can perform many of the learning tasks students once had to complete themselves, what irreplaceable value do universities still offer? Wang Qiong, a professor from the Graduate School of Education at Peking University, argued that a university’s core value lies in “a group of people being present together for a common goal.”

For Chen, what AI cannot replace are experiences that people themselves must feel, take responsibility for, learn from, and make choices about. As students develop, for example, they may confront failed projects, unsuccessful experiments, or rejected papers and, with support from their teachers, search for new approaches. They may also navigate interpersonal friction in genuine collaboration, build trust, and continue asking questions about the meaning of life.

Hu Xiaoyong, director of the Center for Teacher Development and executive deputy director of the Institute of Artificial Intelligence in Education at South China Normal University, argued that if teaching is understood more broadly than the transmission of knowledge and training of skills, the rise of AI does not spell the end of the profession.

Su Dechao, a professor from the School of Philosophy at Wuhan University, noted that AI can provide people with conclusions about values, but values themselves take shape gradually through the feelings, judgments, and inclinations to act that people develop in concrete relationships.

Some Chinese universities have already begun exploring new approaches. Tsinghua University has built an AI general education curriculum system covering five areas and 57 courses, as well as an “AI course matrix” comprising 162 courses, providing multiple levels of entry for students in the humanities, sciences, engineering, medicine, and other fields.

Fudan University has launched more than 100 AI courses and built an integrated undergraduate and graduate “AI-BEST” curriculum system comprising four levels: general foundation, disciplinary core, advanced disciplinary study, and vertical application.

The University of Electronic Science and Technology of China has built a fully integrated “AI-3S” curriculum system, established eight signature courses in AI professional education, and launched an AI micro-major, all aimed at cultivating interdisciplinary “AI +” talent.

Su suggested that universities should not rush to declare a particular model a success, much less regard the number of AI tools used or platforms built as proof that teaching reform is complete.

Citing his own teaching practice, Chen said that after large language models were introduced, the “evidence-based practice design” module in one of his courses was shortened from five or six weeks to two or three. AI took on procedural tasks such as searching, organizing, and calculating, allowing students to devote more of their cognitive resources to selecting topics and critically evaluating evidence.

Chen added that students must submit not just their final results, but also their design approach and the reasoning behind their decisions, explain how they used AI, provide complete interaction records, and justify how they evaluated and selected AI-generated content. The underlying principle, he said, is to encourage appropriate use of AI while requiring transparency and reflection throughout the process.

For Wang, the bottom line of university education is “promoting genuine self-growth and cultivating in students the qualities of seeking truth from facts and pursuing truth and goodness.” Hu, meanwhile, outlined the core competencies universities should cultivate in students for the future: First, an awareness of problems and the ability to ask good questions; second, imagination, interdisciplinary critical thinking, human–AI collaboration skills, and deep humanistic literacy; and third, strong psychological resilience and a capacity for lifelong learning.

Chen argued that the real challenge posed by AI is not whether universities should continue to exist, but what justifies their continued existence. University education, he said, must move from transmitting knowledge to fostering growth, while teachers must move from delivering knowledge to designing learning experiences.