La “legge sull’intelligenza artificiale” dell’UE come punto di arrivo e di partenza dei processi di co-regolazione

di Gabriele Rugani

SOMMARIO:  1. Introduzione: una panoramica dell’iter legis della “legge sull’intelligenza artificiale”. – 2. La “legge sull’intelligenza artificiale” come punto di arrivo del processo di co-regolazione: il dialogo con gli stakeholder avviato fin dal 2018. – 3. La “legge sull’intelligenza artificiale” come punto di partenza di successivi processi di co-regolazione: il ruolo delle regulatory sandboxes. – 4. Riflessioni conclusive: la “legge sull’intelligenza artificiale” come uno strumento di promozione di una specifica modalità di regolamentazione dell’IA.

This contribution intends to focus on a specific aspect of the ‘Artificial Intelligence Act’ proposed by the European Commission in 2021, namely the relationship between this instrument and the issue of co-regulation. First of all, a general overview of the related legislative process will be necessary, starting with the Proposal submitted by the Commission, passing through the positions adopted by the Council and the European Parliament and arriving at the political agreement reached by the two co-legislators. Entering then into the heart of the discussion, it will be possible to highlight how the ‘Artificial Intelligence Act’ represents both a ‘point of arrival’ and a ‘point of departure’ of co-regulation in the field of AI. Point of arrival, because such an instrument is the result of a long and articulated process of stakeholder consultation, which took place within the EU institutions starting in 2018 and intensified after the Commission's 2020 White Paper. A starting point, since, in order to better regulate the delicate subject of AI, it seeks to foster dialogue between national public authorities and stakeholders through additional, even atypical, modes of co-regulation: among these, regulatory sandboxes stand out in particular. In conclusion, it will therefore be possible to carry out some reflections on how the ‘Artificial Intelligence Act’ not only intends to dictate a discipline on AI, but even to promote throughout the Union a specific way of regulating this matter.