AI Safety Connect

Where the world meets to make AI safe

NEW YORK CITY, USA

25.09.2025

How we govern AI now will define our collective future.

Blurred gradient background transitioning from dark blue on the left to bright orange on the right.

As frontier AI systems accelerate toward artificial general intelligence (AGI), the risks associated with these systems pose urgent and unprecedented challenges to international security and stability. 

AI Safety Connect (AISC) is a high-level, invitation-only event series that assists policymakers, entrepreneurs, cutting-edge laboratories, and investors to explore practical safety pathways toward safe and beneficial AI that enables human flourishing. 

AI Safety Connect Objectives

Two people are seated on stage, participating in a panel discussion. The person on the left is an older woman with gray hair, wearing glasses, a tan blazer, and brown pants, holding a notepad. The person on the right is a younger man with short hair, dressed in an off-white suit, holding a microphone, sitting next to a small round table with a water bottle and glasses. Background includes a wooden wall and a small plant.

Our focus is on the interplay between technical AI safety and international coordination. 
Through structured dialogues, we facilitate:

• Situational awareness of frontier AI capabilities and risks.

• Strategic alignment among AI safety institutes, frontier labs, and governments. 

• Practical avenues for coordinated momentum-building for multilateral and international AI safety cooperation.

AI Safety Connect operates as a companion series, strategically following major international events — from the UN General Assembly to the AI Summits — to enable high-quality dialogue and collaboration on effective AI safety governance.

AI Safety Connect UNGA

AI Safety Connect (AISC) is organized by the AI Governance Coordination Project (AIGCP). The UNGA 2025 edition is co-organized with The Future Society, FAR.AI and Mila, and co-hosted with the Permanent Missions to the United Nations from Singapore, Canada, and Brazil, as well as The United Nations Development Programme (UNDP).

With frontier labs openly aiming to achieve AGI, robust Trust & Safety frameworks become the foundational architecture upon which frontier AI risks can be managed and beneficial AI systems can deliver transformative progress toward the Sustainable Development Goals (SDGs). With the global community facing a 35-year low in human development progress, there is an urgent need to ensure that advanced AI systems are developed and deployed safely in ways that close the AI equity gap rather than exacerbate existing inequalities.

To address this critical moment, AI Safety Connect UNGA will convene during the 80th Session of the United Nations General Assembly (UNGA).

Our Partners

A digital logo with the text 'FAR' in orange and a black background.
Text on black-and-white poster reads 'The Future Society' with each phrase in a separate horizontal band.
Mila logo with a hexagonal design to the left of the name

Bringing together high-level stakeholders from government, international organizations, frontier labs, civil society, and academia, AI Safety Connect UNGA seeks to:

  • Advance actionable efforts to define unacceptable uses and behaviors of general-purpose AI systems while establishing shared responsibility mechanisms between public and private sectors.


  • Operationalize frameworks and concrete mechanisms that ensure safety-by-design, equity, and global inclusivity principles are embedded across AI development and deployment cycles.

  • Elevate AI governance on the global diplomatic agenda by supporting multilateral initiatives such as the UN Resolution (A/RES/79/325) to establish an Independent International Scientific Panel on AI and the Report on Urgent AGI Considerations to the Council of Presidents of the UNGA.