Instructors and professors shall transition from being mere knowledge transmitters to learning designers and intelligent guides, while students shall evolve from tool users to human-machine collaborative decision-makers, according to guidelines on the application of generative artificial intelligence in education released by Shanghai's Fudan University on Monday.

The guidelines suggest that teachers, while maintaining classroom leadership, can leverage AI to assist in creating teaching scenarios, designing problem gradients, generating class exercises, and supplementing classroom examples. This allows them to focus more on class discussions, guiding thought processes, and providing personalized feedback—a shift seen as a response to the evolving technological era.

For students, AI is envisioned as a support for autonomous learning, according to the guidelines. Within the permissible scope of course requirements, students can use AI to offload lower cognitive tasks, such as information retrieval and format organization, thereby concentrating on higher-order thinking skills like critical thinking, complex problem-solving, and the development of aesthetic and emotional interaction capabilities.

The comprehensive guidelines, covering areas of classroom innovation, student learning, evaluation feedback, teaching management, faculty development, and student scientific innovation, became the latest case of Chinese universities unveiling documents to set boundaries for the use of generative AI on campus, mainly focusing on academic integrity, usage disclosure, and teachers' and students' responsibilities.

In June 2024, Shanghai-based East China Normal University's School of Communication and Beijing Normal University's School of Journalism and Communication jointly released a guideline on the use of generative AI. It mandated that when students use such tools for assignments, they must mark the AI-generated content, which cannot exceed 20 percent of the full text.

In November, Tsinghua University in Beijing issued guidelines, making it clear that AI should only serve as an auxiliary tool in teaching, with AI use in thesis work needing to be disclosed, and strictly prohibiting academic misconduct and the use of sensitive data to train models.

Regarding Fudan University itself, it issued a regulation on the use of AI tools in writing undergraduate graduation theses in late 2024, detailing areas where such tools are permitted to facilitate research and areas where they are banned. The university said the newly released guidelines align with those regulations, clarifying that AI cannot replace essential academic tasks, such as topic definition, ethical construction, data analysis, and conclusion formation.

"The core issue facing universities with the advent of generative AI is not whether to use it, but how to preserve the essence of education while the technology continues to expand its capabilities," said Lin Wei, head of academic affairs at Fudan University.

In November, an expert committee at the Ministry of Education released the country's first guidelines for the application of generative AI in teaching, emphasizing that teachers must uphold their primary role in education, with AI always being an aid rather than a replacement.

"Such a judgment is also an anchor point for Fudan University in making choices and responses when AI is reshaping knowledge production, learning, and academic ecosystems in universities," Lin said.

Fudan University's new document specifies that course teaching teams can use AI to optimize course module divisions, learning task designs, and activity process arrangements, creating a more cohesive learning loop from lecture, inquiry, and presentation to reflection, and enhancing interactions between teachers and students.

The document further elaborates that for general education courses, AI tools can be used to integrate characteristics from different disciplines to generate diverse cases and scenarios, strengthening real-world relevance and interdisciplinary connections. For specialized courses, AI can quickly capture field frontiers and update teaching materials. For practical courses, AI can design programming and virtual simulation environments, providing real-time operational guidance and feedback, and helping students efficiently master practical skills.

Changes are already visible on the university's campus, with new courses integrating AI with existing content and post-class teacher-student Q&A sessions that provide insights into students' interests and comprehension levels, making classroom content more targeted.

Zhang Hao, an associate professor at Fudan University's College of Future Information Technology, has introduced a new course on semiconductor device physics by systematically redesigning the curriculum using AI-based methodologies, which has been warmly received by students.

Wang Yanjin, who teaches at Fudan University's School of Stomatology, mentioned that the AI tools created for the course "Oral Histology and Pathology" include AI virtual patients, allowing students to engage with patients of different personalities and simulate real-world professional scenarios.

"Through such AI experiments, students' interest in the entire course increases, leading to more explorations and thoughts in areas like pathology," she said.