Transcript Videomessage
Ladies and gentlemen, dear friends, it’s a pleasure to join you, even if only on screen, for the 20th anniversary of the European Summer School on Internet Governance.
Let me begin with where we stand. In the Global Digital Compact, member states recognize the Internet as a critical global facility. They reaffirm the multistakeholder model of Internet governance along with the outcomes of Geneva and Tunis. That’s significant at a time when the global governance landscape is shifting significantly.
But commitments need to be implemented by people who understand the substance deeply enough to carry them forward. And many of the commitments made by member states rely on contributions from across the private sector, civil society, and the technical community. That’s why capacity building as done by the summer school is at the centre of making Internet governance work. And this is where our conversation today naturally turns from governing the network that connects us to governing the intelligence now running on top of it.
Internet governance and AI governance are converging. They are not identical. There are real differences in how these technologies are built, deployed, and controlled. But there would be no AI without the Internet. And the principles we developed for Internet governance, including openness, interoperability, and meaningful multistakeholder participation, can serve as a source of inspiration for new concepts in AI governance.
The principles of Tunis still ring true for AI.
However, we live in a completely different technological and geopolitical era. It took the Internet a decade to reach 1 billion people. For AI conversational models, it took 3 years. The main technology companies propelling the AI boom are well established behemoths with billions of users and annual revenues exceeding hundred billion dollars and value comparable to the GDP of a G7 country.
This concentration of power and capacity in the hands of a few private actors is something we have to contend with as we “refine” the multistakeholder model.
Enhanced cooperation is another key term long debated and contested by delegates and experts. This cooperation was conceived in the words of the Tunis agenda to quote “enable governments on an equal footing to carry out their roles and responsibilities in international public policy issues pertaining to the Internet” unquote. As the secretary general noted in his report on innovative financing options for AI capacity building, we should ask what cooperation and enablement mean for countries that still face foundational capacity gaps in AI.
So the question arises, how can we leverage the multistakeholder and private sector led nature of digital technologies to prevent the data and AI divide from getting even wider? The WSIS process offers lessons and cautionary tales.
Some of these lessons were carefully reflected in the Global Digital Compact which represents now the compass for the international community to address AI and emerging challenges. So what are these?
First, we need space for discussion. The IGF emerged at the tail end of 1990s UN summit diplomacy and consolidated a vibrant community around it to discuss the most pressing issues in Internet governance. And now the Global Dialogue on AI governance which met for the first time in Geneva this past July can offer such a space and ensure that all stakeholders including the over 100 countries excluded from main mechanisms on AI governance are part of the conversation. We look forward to seeing you in New York on 3rd and 4th May 2027 for the second dialogue after the successful hosting of the first one in Geneva.
The second lesson is that we need to address capacity development early and partner with international financial institutions, philanthropies and the private sector.
The proposal of the Secretary General for a global fund for AI capacity building is still being discussed. In the meantime, a network of capacity development centers across all five continents of the globe has come together. This is a testament to the enduring role of the technical community and academia in closing digital divides 20 years after Tunis.
The third lesson that we need to carry over into the AI era is that we need expert advice on the risks and opportunities of digital and emerging technologies. And this is why the General Assembly has created an independent international scientific panel on AI which issued its first report this July.
To conclude, this community, your community knows the terrain. You built the practices that kept the Internet open and interoperable. Bring that wisdom to the governance of AI. How people find information, how we build infrastructure, how we keep users safe.
This school, the series of summer schools around the world and the IGF ecosystem turn debates into practice. So keep that working. Surface the early signals, test ideas across regions, and help translate good concepts into guard rails and capacity where they are needed the most. 20 years on, the task is the same. Build bridges across disciplines, sectors and borders so that the Internet and AI widen opportunity and not divides. Thank you.