This first edition of this event was a resounding success, with engaging discussions and insightful panels. In addition to the more than 100 in-person attendees, more than 500 persons connected online from around the world for the high-level public segment of the event.
The spotlight of the event was on AI supervision, AI and democracy, and Synthetic Content. This event brought to fruition UNESCO’s efforts to implement the 2021 Recommendation on the Ethics of AI, including comprehensive diagnostic tools, regional roundtables for peer learning, and experts from around the world. 
In the context of this event, the panel titled Global Landscape of AI Governance: the Role of Supervising Bodies, gathered different supervisory authorities and regulatory bodies working on AI supervision for their expert views and insights.
The panel began with a presentation by the project coordinator, who outlined the objectives and deliverables of the project. Additionally, preliminary results of a survey conducted from December 2023 to May 2024 to supervisory authorities members of the European AI Working Group for Competent Authorities were presented. The survey results highlighted the importance of the role in the implementation of legal framework by these authorities, their challenges in learning and adapting quickly to enforce new rules against the background of a fast-paced development of technology, and finally, in this context, the need to support and reinforce peer-to-peer learning between authorities.
This presentation was followed by a high-level discussion between UNESCO Assistant-Director General for Social and Human Sciences, Gabriela Ramos, the Director of the European Commission Directorate-General for Structural Reform Support(DG REFORM) Nathalie Berger, and the Inspector General of the Dutch Authority for Digital Infrastructure (RDI), Angeline van Djik. This discussion allowed for the unveiling of two deliverables from the project : first the Global Landscape of AI Supervision Report and second, the first training material. The training material was piloted successfully in May with the European AI Working Group for Competent Authorities.
The high-level discussion was followed by a panel discussion on the global landscape of AI supervision through the lens of seven experienced regulators. This discussion focused on the collaborative efforts needed to leverage resources and knowledge and explored the growing importance of international networks in facilitating cooperation and capacity building for supervising authorities. The discussion also brought to light several key issues facing global cooperation in regulating AI. The key takeaways from the panel can be summarized as follows:
Experts emphasized the importance of establishing common ground and shared understanding among different authorities.
One of the primary concerns raised was the need for a unified definition of terms, particularly when it comes to assessing risk. Panelists agreed that fostering international cooperation hinges on developing a shared vocabulary and conceptual framework.
The importance of empirical data was also underscored. Experts suggested that gathering and analyzing concrete information could lead to a deeper understanding of common issues and potential solutions in AI supervision.
Defining the scope of cooperation emerged as another crucial point. Panelists debated whether the focus should be on interoperability of regulations and standards, or if it should extend to other areas of collaboration. Technical cooperation, such as joint testing initiatives, was proposed as a potential avenue for international collaboration in AI supervision.
A recurring theme throughout the discussions was the delicate balance between innovation and enforcement. Panelists acknowledged that effective enforcement requires a deep dive into technical details, which in turn enhances the overall understanding of AI systems and their implications.
Overall, this panel highlighted the complex landscape of AI supervision on a global scale. As nations grapple with the rapid advancement of AI technologies, finding common ground and establishing effective international cooperation frameworks remains a pressing challenge.
Finally, the moderator of the panel also opened the floor to collect comments and questions from the public. Questions on the need to focus on the interoperability of approaches, sustainability, and the importance of small communities and minority languages were raised by the public and further discussed by the panelists.
The project Supervising AI by Competent Authorities aims to respond to those needs in the context of the EU, as the project is funded by DG REFORM. The panel confirmed that beyond the territorial scope of the project, there is a need and interest to create a network through which different authorities could facilitate the exchange of good practices given the cross-border nature of AI. In the context of the project, we will dive into this conversation in the immediate future seeking to reaffirm the importance of supporting capacity building between supervisory authorities in the EU and beyond.
Speakers of this session:
Moderator:
- Huub Janssen, Chair of Dutch and of the European Working Group of Competent Authorities on AI, and AI Manager at the Dutch Authority for Digital Infrastructure
Presentation of the project:
- Angelica Fernandez, Ethics of AI unit, Social and Human SciencesSector, UNESCO
High-level panel on AI supervision:
- Gabriela Ramos,Assistant Director-General, Social and Human Sciences,UNESCO
- Angeline van Dijk,Inspector General, Dutch Authority for DigitalInfrastructure (RDI) (online)
- Nathalie Berger, Director,Directorate-General for Structural Reform Support(DG REFORM) - European Commission (online)
Panel on Global Supervision Landscape and Q&ASetting the Scene:
- Elham Tabassi,Chief Technology Officer, US AI Safety Institute
- Alexandra Schultz,Advisor, Norway Digitalisation Agency
- Jordi Ascensi Sala, Head of Technology, Andorra Research Innovation (AR+I)
- Paloma Jesús Herrera Carpintero, Advisor to the Minister, Ministry of Science of Chile
- Lee Wan Sie, Director of Development of Data-Driven Tech, Singapore Infocomm Media Development Authority (IMDA)
