In this 30-minute video, Thibault Schrepel, the Network Law Review’s creator, converses with Axel Voss, Member of the European Parliament and rapporteur on the AI Act. Together, they explore the main lines, approach, philosophy, and ambition of the forthcoming Artificial Intelligence Act.
In a nutshell, here is what to make of the conversation. We addressed the following issues:
➝ Is the European Commission’s sectorial approach too strict?
➝ Will the AI Act be manageable—as it stands—for developers?
➝ Should expert systems be covered?
➝ Should we alter the requirement to use “free of errors and complete” datasets?
➝ How to make the AI Act more than a goldmine for the compliance industry?
➝ How can we minimize regulatory capture?
➝ Will the AI Act be compatible with GDPR?
➝ Will the AI Act help European innovation versus US innovation?
➝ Will the AI Act put small and medium companies at a disadvantage?
Axel Voss made several announcements:
➝ There are discussions to create a European “AI Agency”
➝ MEP know, already, that not enough experts will be found to certify AI systems, so the Parliament may introduce other options
➝ In his views, GDPR should be amended, but there is “no political courage or will to do so right now” = not likely to happen
And finally, he shared his position on several subjects:
➝ The current definition of “high-risk systems” is “nonsense” and should be changed
➝ GDPR interpretation by national agencies is too board and should be constrained
➝ The AI Act should not aim at “consumer protection,” but “competition”