WIC holds symposium on frontier AI development and safety risks

The World Internet Conference Specialized Committee on Artificial Intelligence (WIC SC on AI) held a symposium under the theme  “Frontier AI Development and Its Safety Risks” in Beijing on March 20.

The event was part of the committee’s AI Safety and Governance Program, which was co-organized by the Beijing Institute of AI Safety and Governance.

Members of the specialized committee and more than 50 AI experts and academics from around the world gathered for a discussion, both online and offline, about the framework of a thematic report on risk understanding and coordinated governance of frontier AI they are working on.

The meeting was chaired by Zeng Yi, co-chair of the WIC SC on AI, co-lead of the AI Safety and Governance Program, and dean of the Beijing Institute of AI Safety and Governance. He presented the 2026 work plan for the program and the draft report framework. 

The meeting was attended by advisors of the committee - including Gong Ke, executive director of Chinese Institute of the New Generation Artificial Intelligence Development Strategies, and Xue Lan, professor and dean of Schwarzman College at Tsinghua University, as well as vice chairs - including Zhang Yong, vice president of QiAnXin Group, and Pan Jianfeng, chief scientist of 360 Group and CTO of 360 Security Technology. They were invited to deliver keynote remarks focusing on the meeting theme and report framework.  

Seán Ó hÉigeartaigh, co-lead of the AI Safety and Governance Program and research professor of the Centre for the Future of Intelligence at University of Cambridge, joined a representative of the WIC, in delivering concluding remarks.

With focus on the new safety challenges arising from rapid advances in frontier AI technologies and their accelerated application, the upcoming report would analyze global consensus and divergence in AI risk understanding, and assess the challenges facing existing governance frameworks in terms of adaptability and coordination, according to Zeng.

It would, he added, explore ways to strengthen multi-stakeholder collaboration and enhance governance effectiveness, offering research-based guidance for the international community.

A wide range of experts and academics had in-depth exchange of ideas on the draft report framework, including those from WIPO, OECD, UNESCO, Smart Africa, Chartered Institute of Arbitrators, the University of Muenster, the Chinese University of Hong Kong, Peking University, Nankai University and China Academy of Information and Communications Technology.

Also involved in the discussion were those from companies such as 360 Security Technology Inc., Alibaba Cloud, China Mobile, DAS-Security, Huawei, Lenovo Group and MiniMax.

They emphasized the importance of systematically identifying and assessing frontier AI risks, analyzing governance gaps and promoting collaborative multi-stakeholder approaches to governance. 

A WIC representative said the drafting team would sort through and incorporate their suggestions to refine the report framework, and continue drafting efforts. Based on this joint research, the WIC plans to produce a series of annual reports on frontier AI safety and governance. 

The series will track AI developments and the safety and governance challenges that come along, assess critical risks and promote the improvement of governance practice, to boost global consensus for safe, reliable and responsible frontier AI development. 

The findings are scheduled for official release at the 2026 WIC Wuzhen Summit.