AI is having a highly disruptive impact at universities. University leaders gathered to discuss how best to harness the technology.Credit: Eugene Mymrin/Moment/Getty
The leaders of global universities who recently gathered in Kyoto, Japan, concur on one thing: artificial intelligence (AI) is a highly disruptive technology with massive potential to accelerate research and enhance education. But it is also raising all kinds of challenging questions and ethical concerns, and nowhere is this double-edged sword more evident than on university campuses.
“AI is transforming the culture and values of universities,” Kensuke Fukushi, director of the Institute for Future Initiatives at the University of Tokyo noted at the gathering. “It differs from earlier technologies in that it challenges the very foundation of how knowledge is produced and validated,” he said.
This reality is prompting critical questions about the future of creativity, originality and trust. As universities are the birthplaces of many of the foundational breakthroughs behind AI, they are both developers and adopters, and have a key role to play in ensuring its responsible, inclusive and ethical use.
Teruo Fujii, president of the University of Tokyo.Credit: Mika Hashimoto
These issues were a major topic of discussion during the event, which was a session titled ‘AI and the Future University: Strategic Leadership, Culture, Diversity and Global Equity’.
This was organized by the International Alliance of Research Universities (IARU) as part of the 22nd Annual Meeting of the Science and Technology in Society (STS) forum held in Kyoto in October. The session considered how strategic university leadership can integrate AI while maintaining academic integrity, cultural pluralism and social trust.
The STS forum is an annual conference that fosters dialogue among hundreds of global leaders in science, technology and policy. It hosted the session organized by IARU, which is a consortium of 11 leading research-intensive universities working together to address global challenges.
The session brought together presidents and senior executives from leading universities across five nations. It was co-chaired by Teruo Fujii, president of the University of Tokyo, and Mosa Moshabela, vice chancellor and principal of the University of Cape Town in South Africa.
Mosa Moshabela, vice chancellor and principal of the University of Cape Town.Credit: Mika Hashimoto
One highlight was a panel discussion between Qihuang Gong, president of Peking University in China; Gunther Dissertori, rector of ETH Zurich in Switzerland; and Bin Liu, deputy president of the National University of Singapore. The conversation was moderated by the session co-chairs.
Learning curve
Providing commentary at the session’s conclusion, Nick Campbell, vice president for academic affairs at Nature’s publisher Springer Nature, compared the challenge of adapting to rapidly developing AI technologies to building an aeroplane in midair, with little more than improvisation to rely on.
One approach to navigating this at the University of Tokyo, Fujii said during the discussion, was the release of guidelines around ChatGPT use soon after its launch in 2022. “We encouraged students not to obtain answers directly from AI,” he explained, with the university instead asking them to use AI to research information that helped them evaluate their own answers.
“Education is more than collecting knowledge; it’s important to learn from students’ field experience itself,” Fujii added.
ETH Zurich’s Dissertori pointed out that AI will likely completely overhaul how education operates.
“Educators will have to transform themselves from holders and transmitters of knowledge into coaches that train people how to use [AI] in a critical and ethical way,” he said. “I think that is the biggest challenge we’re facing now, and as universities, we can take the lead.”
Stewarding AI
All the panelists agreed that universities have a key role to play in ensuring that AI is used for the public good and that it doesn’t deepen global inequalities.
Peking University’s Gong noted that a major challenge is ensuring accountability and transparency. Dissertori agreed, emphasizing that transparency, responsibility and fairness were the three guiding principles for ETH Zurich when it put together its current AI strategy.
Bin Liu, deputy president of the National University of Singapore.Credit: Mika Hashimoto
A key concern that prompted the meeting of university leaders was ensuring that AI promotes global equity and diversity — a principle reflected in the broad mix of panelists and commentators. Moshabela noted that there is a danger, for example, of African institutions missing out on the advantages of AI because of limited investment and infrastructure to support adoption.
“We need to ensure that Africa does not get left behind,” he urged delegates.
Fujii noted that the dominance of the English language in material used to train large language models (LLMs), combined with the need for vast amounts of data, is causing many existing western-developed AIs to become heavily English language- and culture-centric. “LLMs today are mainly based on English-language data,” he told attendees. “We need to find a way to incorporate multiple linguistic and cultural characteristics into AI.”
Efforts are underway to make AI more diverse. Liu mentioned a US$100 million initiative in Singapore that is incorporating data from different languages into a single open-source LLM.
“This model really leverages the cultural and linguistic diversity in southeast Asia,” Liu said. “If you’re interested in anything related to the language or culture, I’m very confident you will find insight into it via this AI platform.”
Gunther Dissertori, rector of ETH Zurich.Credit: Mika Hashimoto
Similarly, Dissertori described a Swiss LLM trained on more than 1,000 global languages.
Moshabela noted that “AI can also promote diversity by helping overcome language barriers, and that the University of Cape Town is exploring the use of AI in classrooms to improve the experience of students struggling with the language being used”.
Ongoing conversation
A common thread throughout the discussions was the importance of considering long-term impacts. For example, Liu pointed out that short-term gains in efficiency could lead to people losing their jobs. “We have to strike a balance between the benefits and the issues that AI causes,” she said.
Reflecting on the event after guiding these discussions, Fukushi — who was the facilitator of this session — said it will be important not to try to squeeze every solution into the same mould. “The exchange reminded me that AI strategies must be context-sensitive and inclusive, rather than one-size-fits-all,” he said.
This was one of the major global, presidential-level dialogues linking AI with university leadership, culture and global equity. It also marked the launch of a long-term collaboration between IARU and the STS Forum on an AI Leadership Roadmap for Universities.
Qihuang Gong, president of Peking University.Credit: Mika Hashimoto
“Further discussions are ongoing, and AI is likely to remain a hot topic for years to come”, said Fujii.
In the wake of the gathering, he added, “I’d like to see more dialogue on how AI can promote global equity — especially in supporting researchers in the Global South — and how academic reward systems might evolve to fairly recognize AI-assisted contributions.”
The conversations sparked at the STS forum signal the beginning of a shared journey to ensure that the age of AI advances both knowledge and humanity. For universities, the task ahead is clear: to oversee AI not merely as technology, but as a moral and educational responsibility to shape the future of knowledge itself.
