Hello.
I would like to once again express my gratitude to our co-hosts, the Republic of Korea, and to Minister Lee for bringing us together for this extremely important event.
Your dedication and leadership have made this gathering possible, just six months after we met at Bletchley Park last November.
It is truly moving to see so many familiar faces around the table again and to reflect on the extraordinary progress we have made since our last meeting.
We have a shared mission: to understand the risks in order to seize the great opportunities together across public services, business and society as a whole.
When I spoke at the UK AI Safety Summit, I talked about how governments quickly came together to ratify the Montreal Protocol and tackle the existential threat of a hole in the ozone layer. I hoped that this example would serve as a benchmark for what we can achieve together through these summits. And I'm pleased that we've moved even faster on the global challenge of AI safety.
At Bletchley we announced the launch of the UK AI Safety Institute, the world's first government-backed body dedicated to advanced AI safety for the public good.
Since then, the Institute has made significant progress using state-of-the-art tools to conduct pre- and post-deployment testing of advanced Frontier models.
Yesterday, leaders agreed to bring these laboratories into a global network to show that the “Bletchley effect” is actually happening.
Last night, leading companies around the world signed the Frontier AI Safety Commitment, marking a historic moment by pledging to make AI safer and refrain from releasing new models if the risk is too high. I welcomed you.
And last week, we published our Interim International AI Safety Report on Advanced AI Safety, bringing together the latest scientific research to inform policymaking based on solid evidence.
Together, we accomplished all this within six months, demonstrating the power of global collaboration and action-oriented focus.
But my message today is that we must not be complacent. As the pace of AI development accelerates, we must match our own efforts and actions to understand risks and seize opportunities.
Phase 2, from Seoul to France, will push the boundaries of early science: frontier AI testing and evaluation. At the same time, attention should be given to risk mitigation beyond these models. Ensuring that society as a whole is resilient to the risks posed by AI.
This is where the concept of system safety really comes into play, where AI safety mechanisms are built into the systems of society, not just the AI systems themselves.
The UK has already started investing in systemic safety research and we want to deepen global collaboration on this important issue.
In 1988, the year the Seoul Olympics were held, Isaac Asimov said, “Science gathers knowledge faster than society gathers wisdom.”
It is our responsibility to ensure that human wisdom keeps pace with human knowledge as we navigate the most complex technology humanity has ever created.
I look forward to engaging in detailed discussions and agreeing on concrete, tangible actions to move forward with the next steps in global AI security policy.
International collaboration between AI safety agencies and their counterparts is an absolutely essential mission for our shared success, and today's showcase is an opportunity to learn from each other and build a sustainable and productive This provides an opportunity to build a foundation for further cooperation.
Now, I would like to hear from Minister Li.
