AI Safety Connect at UNGA 2025
United Nations Headquarters, Delegates Dining Room
September 25, 2025
AI Safety Connect at UNGA 2025
Advancing International Collaboration on the Governance of Advanced AI Systems
September 25, 2025 | United Nations Headquarters, Delegates Dining Room | 6:00PM to 9:00PM
EVENT
Building on the success of the inaugural AI Safety Connect event at the 2025 AI Action Summit in Paris, this high-level convening brought together over 135 senior diplomats and AI practitioners, seizing a critical moment to elevate AI safety on the global diplomatic agenda, this high-level convening brought AI governance experts, frontier lab leaders, and policy scholars at the United Nations Headquarters, Delegates Dining Room during General Assembly week.
AI Safety Connect (AISC) is organized by the AI Governance Coordination Project (AIGCP). This "at UNGA 2025" edition was co-organized with The Future Society, FAR.AI and Mila, and co-hosted with the Missions to the United Nations from Singapore, Canada, and Brazil, The United Nations Development Programme (UNDP).
Opening Remarks From Ministers and Senior Officials Set the Stage
M.C. Mr. Cyrus Hodes
Founder AI Safety Connect
H.E. Josephine Teo,
Minister for Digital Development and Information of Singapore
USG. Amandeep Singh Gill,
Under-Secretary-General, United Nations
H.E. Elina Valtonen,
Minister of Foreign Affairs of Finland
H.E. Minyoung Han,
Director-General for Climate Change, Energy, Environment and Scientific Affairs, Republic of Korea
Mr. Robert Opp,
Chief Digital Officer, United Nations Development
Programme ( UNDP)
H.E. S. Krishnan,
Secretary, Ministry of Electronics and Information Technology, Government of India
Ms. Sulyna Abdullah,
Representing the ITU Secretary-General
Keynote addresses
Prof. Yoshua Bengio (lead author of the 2025 International AI Safety Report and a pioneer of artificial neural networks and deep learning) and Prof. Anthony Aguirre (Executive Director, Future of Life Institute) provided scientific grounding for the evening's discussions.
“The bad news is, we simply don't know how to design AI that will assuredly not hurt people, that will be deployed and be used in a way that is in line with humanity’s interests [...] We must act now to mitigate all of these risks […]”
-Prof. Yoshua Bengio
"There is a tendency to consider the development of AI like a train on a single track and our only option is to decide how fast or slow we steam ahead, and I think this is not right. AI development as with any technological development is a proliferation of paths and [...] we can and must choose [...] not to go down paths that are widely agreed to present threats to security of the world's nations and people [...] we must instead choose paths toward pro-human AI tools that empower us all."
-Prof. Anthony Aguirre
A Fireside Chat on Red Lines to Prevent Unacceptable AI Risks
Co-moderated by Niki Iliadis (Director, Global AI Governance, The Future Society) and Charbel-Raphaël Segerie (Executive Director, Center for AI Security), announced the launch of the "Global Call for Red Lines for AI" campaign and explored practical pathways toward international coordination.
Panel 1: Safety on the Frontier of AI
Moderated by Dr. Adam Gleave (CEO, FAR.AI), featured a conversation with Dr. Chris Meserole (Executive Director, Frontier Model Forum), Natasha Crampton (VP, Chief Responsible AI Officer, Microsoft), and Coco Zhou (Senior VP and Partner, UCloud) on preparing for a world with AGI and implementing robust frontier AI safety frameworks.
Panel 2: The Role of AI Safety Institutes
Moderated by Karine Perset (Head, Division of AI and Emerging Technologies, OECD), convened Dr. Xiao Qian (Vice Dean, Institute for AI International Governance, Tsinghua University), Dr. Deval Pandya (VP, Vector Institute and Advisor, Canadian AI Safety Institute), and Wan Sie Lee (Director, Infocomm Media Development Authority of Singapore) to discuss international cooperation mechanisms and best practices.
Closing Remarks
Co-founders Cyrus Hodes and Nicolas Miailhe closed the event with gratitude to our speakers' contribution to AI Safety Connect and to global AI Safety.
The program concluded with a networking reception that facilitated substantive bilateral conversations between senior diplomats and AI safety experts—fostering the relationships and communication channels essential for coordinated action as AI capabilities continue to advance.
Photo Gallery
Attendees can access the gallery using the code provided in their "Thank you for attending" email. Please use the same email address as provided during registration.
