The Ambiguous Risk-Based Approach of the Artificial Intelligence Act: Links and Discrepancies with Other Union Strategies

The AI Act regulation proposal adopts a risk-based approach to the regulation of artificial intelligence systems. As a matter of fact, the risk-based approach has become more typical of Union strategies with respect to digital policies. However, the way such an approach has been declined varies greatly: most notably, whereas the GDPR and, to a limited extent, the DSA regulation proposal adopt a bottom-up perspective, the AI Act rather reflects a top-down scheme, where the task of risk assessment is kept within the hands of the legislator. This position paper aims at highlighting the common features, as well as the differences, between the various legal acts discussed: in particular, by considering (optimal) proportionality and due diligence as a characterizing features of the risk-based approach, the goal is to understand whether the AI Act does indeed reflect the typical principles of this developing legal model. Although noting that the role of due diligence is feebler within the regulation proposal, we argue that the central common point is represented by the (constitutionally relevant) goal of proportionality.