About AI Safety Connect
How it started
2024 | PARIS AI ACTION SUMMIT
Described as “the most substantive side event" by participants at the 2024 Paris AI Action Summit, AI Safety Connect launched in 2024 with the generous support of the Mohammed Bin Rashid School of Government (MBRSG) and the Future of Life Institute.
How we operate
By focusing on the interplay between technical AI safety and international coordination, AI Safety Connect brings academics, researchers, entrepreneurs, frontier labs, policymakers, and investors together to seriously explore the full range of AI development scenarios and timelines, ensuring that all stakeholders can contribute to shaping safe and beneficial AI governance frameworks.
We operate as a companion series, strategically following major international events such as the UN General Assembly, AI Summits, and key academic conferences to build coordinated momentum and opportunities for high-quality dialogue and collaboration on effective AI safety governance.
Impact
2025 | PARIS AI ACTION SUMMIT
Our inaugural event at the Paris AI Action Summit in February 2025 successfully convened over 150 distinguished participants, including:
Turing Award winner Yoshua Bengio and Professor Stuart Russell
Representatives from OECD.AI/GPAI, UNICRI, Microsoft, Government of India, Government of Singapore, Frontier Model Forum
Senior coverage from Axios, The Financial Times.

Why AI Safety Connect is Needed
The promise of transformative AI systems creates profound uncertainty: how do we build future-proof technical and social governance systems for a technology that is fast and profoundly reshaping economies, societies, and the very conditions of post-WWII international security?
No one lab, state, local government, trade union, or any other relevant stakeholder can face this challenge alone. By bringing together technical experts, frontier AI companies, policymakers, academics, civil society representatives, industrial players, media, and investors, AI Safety Connect seeks to weave a fabric of cooperation towards action.
We operate on the following assumptions about transformative AI systems:
Imminent possibility: Decision-makers must grapple with considerable uncertainty around AGI development timelines. Transformative systems may emerge within quarters rather than decades, making it essential that current leadership help shift away from dangerous racing dynamics toward responsible, coordinated approaches for safe and beneficial AI.
Global cooperation imperative: It is vital and urgent to forge connections between decision-makers across east-west and north-south perspectives, facilitating meaningful exchanges across traditional silos to build the relationships and communication channels necessary for effective coordination if/when AI capabilities spike to dangerous levels.
Democratic participation in AI's trajectory: Transformative AI systems are fundamentally reshaping economies, societies, and human agency. Policymakers and the public must be informed participants who can meaningfully influence the direction of AI development, not passive recipients of technological change.