On December 2, 2024, the International AI Cooperation and Governance Forum 2024, jointly hosted by Tsinghua University, the National University of Singapore (NUS), and The Hong Kong University of Science and Technology (HKUST), opened at NUS. With the theme of “International Cooperation on AI Governance,” the Forum gathered over 60 world-renowned artificial intelligence (AI) experts, scholars, industry leaders, government representatives, and international organization delegates to explore the establishment of an open, inclusive, and mutually beneficial international AI governance framework.
Officiating guests of the Forum were Lew Chuen Hong, Chief Executive of Singapore’s Infocomm Media Development Authority; Tshilidzi Marwala, Rector of the United Nations University (UNU) and Under-Secretary-General of the United Nations (UN), Yang Bin, Vice President of Tsinghua University; Simon Chesterman, Vice Provost at NUS; and Guo Yike, Provost of HKUST.
In his role as chair of the opening ceremony, Simon Chesterman noted that this Forum aims to find the path of international AI cooperation, particularly the two core challenges facing AI governance. Firstly, the governance frameworks are often limited to the perspectives of a few countries. Secondly, discussions on governance often remain superficial at the level of abstract principles without in-depth and detailed analysis.
Yang Bin welcomed all the participants and said in his remarks, “Tsinghua University seeks to explore solutions from experts and scholars worldwide to address the technological limitations of AI development in the international scientific community through inclusive and open scientific research on AI, and to provide scientific and reliable intellectual support for ‘good governance’ and ‘good use’ of AI.”
In his welcome remarks, Guo Yike stressed the importance of aligning AI governance with ethics and emphasized the combination of technology and policy efforts. He welcomed the participation of the global academic community and practitioners and highlighted the concerted efforts of the entire international community in the research of this field.
Lew Chuen Hong said in his opening remarks that ASEAN countries could lead in inclusive AI governance through open programs and diverse perspectives. “There are two key aspects of AI governance. First, we should translate governance principles into actions and make them operable, traceable, and executable, which requires joint efforts. Second, we must integrate the political and technological communities to ensure that AI progress is rooted in national policies and a sound scientific foundation,” he concluded.
Tshilidzi Marwala extended greetings to all participants on behalf of UNU. According to Marwala, confronted with the opportunities and challenges brought by the rapid development of AI, the UN is working to improve AI governance to build a global digital governance framework that is inclusive, sustainable, and fair. He encouraged global collaboration to create a more inclusive, sustainable, and secure future.
The keynote speech session was moderated by Gong Ke, Executive Director of the Chinese Institute of New Generation Artificial Intelligence Development Strategies and Director of the Xinchuang Haihe Laboratory. Keynote speeches were delivered by Martin Hellman, Turing Award winner and Professor Emeritus of Electrical Engineering at Stanford University; Xue Lan, Dean of the Institute for AI International Governance (I-AIIG), Tsinghua University; Dame Wendy Hall, Regius Professor of Computer Science at the University of Southampton; and Guo Yike, Chair Professor and Provost of HKUST.
Gong Ke noted that in the past two years, the world has witnessed the rapid progress of advanced AI technologies, such as ChatGPT and Sora, alongside rapid developments in AI governance. “AI should be aligned with human society in common sense and sustainable development, and we should broaden international dialogue to develop AI in a safe, responsible, and trustworthy manner,” he emphasized.
Martin Hellman delivered a keynote speech titled “Open Science,” indicating that science knows no borders and that we must strengthen cooperation in today’s world, where technology and national security are increasingly intertwined. He stressed that to avoid unprecedented disasters caused by technological progress, we need to adopt a new way of thinking and recognize that national security increasingly depends on the global security of all countries. “Drawing from historical precedents, I encourage international scientists to collaborate to prevent global disasters and to jointly pursue human well-being and progress,” he concluded.
Xue Lan observed that AI is progressing at unprecedented speed in 2024. To address its governance challenges, multiple stakeholders have actively participated in international AI governance initiatives and made significant achievements. Given that different stakeholders have varying advantages and disadvantages in legitimacy, inclusiveness, effectiveness, and scientificity of governance, it is imperative to establish a coordinated and effective international governance system to give full play to respective advantages to address the rapidly developing AI technology. At the end of the speech, Xue issued a report on Promoting AI Governance under the Principles of Extensive Consultation and Joint Contribution, which was led and drafted by I-AIIG. The report provides recommendations for establishing an inclusive global AI governance system.
Wendy Hall mentioned that the UK government announced a policy emphasizing innovative and iterative approaches to AI regulations on March 29, 2023, and invested £31 million in an AI safety research program led by the University of Southampton. This trend reflects growing concerns about deepfake and content copyright incurred by the rapid progress of large language models (LLMs) like ChatGPT, particularly regarding the lack of diversity caused by AI. “We should further enhance cooperation and attention and ensure the safety of AI through in-depth international cooperation,” she stressed.
In his keynote speech, Guo Yike discussed the OpenAI o1 paradigm of rational LLMs, value-based reasoning alignment, training of LLMs for legal reasoning, and the application of LLMs in legal technologies through LexiHK. Guo pointed out that value-based reasoning is the core of AI alignment, and building AI systems requires going beyond the interests of individuals or organizations in a legal and reasonable manner.
Other highlights of the Forum were the panel discussions chaired by the Forum’s supporting partners - I-AIIG, AI Singapore, AI Verify Foundation, Hong Kong Generative AI Research and Development Center, and Concordia AI. Discussion topics included International Cooperation on AI Governance, AI Safety, AI Ethics and Governance, AI and the Future of Work and Education, AI Industry Development and Governance, The Public Sector and AI, AI for Law, and AI for Sustainable Development.
The Forum also gained support from organizations worldwide, including UNDP China, UNU Macau, and Infocomm Media Development Authority. Scholars from renowned institutions, including Stanford University, University of Southampton, Technical University of Munich, University of Melbourne, University of Surrey, and Sungkyunkwan University, as well as senior executives from well-known scientific and technological enterprises such as Microsoft, IBM, Tencent, and LinkedIn, joined the Forum.
