In the Breakout Session on AI and Governance at the EU High-Level Conference on AI - From Ambition to Action last September, we took a closer look at the governance framework of the European Commission’s Proposal for a regulation on Artificial Intelligence (AI Act). The session saw stakeholders who are directly involved in different stages of the governance process come together to discuss how to move towards a successful implementation of the proposed legal framework. The discussion thrived on the exchange of their very different perspectives and offered several valuable lessons.

Firstly, much could be learned from the implementation of the General Data Protection Regulation (GDPR) and applied to the implementation of the AI Act.

Since the European Data Protection Supervisor (EDPS) will be a member of the newly set up European Artificial Intelligence Board tasked with facilitating a harmonized application on the EU level, valuable experience from the enforcement of a similarly complex horizontal piece of legislation, the GDPR can be levered. For instance, it was mentioned that there is a strong need for a single point of contact and exchange between the relevant national authorities on an EU level.

Secondly, the implementation of the AI Act requires adequate expertise and capacity.

On the day-to-day part of the implementation, independent notified bodies will assess individual compliance, and national authorities will supervise the application of the rules and carry out market surveillance activities. Panel members endorsed the approach to governance taken by the AI Act and report that authorities and notified bodies are already building up resources to prepare. However, while preparations are ongoing, a key point of concern and takeaway is the current shortage of expertise and the need for educational initiative and capacity building.

Thirdly, the balance between protecting human rights and facilitating AI innovation has to be struck.

The proposed AI Act pursues the twin objectives of addressing the risks associated with specific AI applications and promoting the uptake of AI. The aim of the initiative is a human-centric approach to AI. At the same time, it is important to create an ecosystem that fosters innovation.While some panelists highlightedregulation should be put on the forefront in order to ensure a high level of fundamental rights protection, others stressed that small-scale providers and start-ups are drivers of innovation and further regulatory intervention could put them at a disadvantage to larger companies. These different perspectives demonstrate the need to strike the right balance between effectively ensuring compliance while not overburdening innovators. We thank our panelists for their valuable contributions and the committed audience for their numerous interesting questions and impulses.

A summary report from all the sessions of the event will be published soon on the AI Alliance while a series of posts, will address the questions received during the two days of the event. In the meantime, we invite you to watch the recordings of the session and post possible questions/comments in the feed below.