The European Parliament has recently adopted several resolutions on artificial intelligence, including the Civil Liability Regime and Artificial Intelligence of 20 October 2020. The latter, which was followed by the proposal for a Regulation of 21st April 2021, introduces different types of artificial intelligence systems and differentiates their liability regime: it distinguishes, in fact, high-risk artificial intelligence systems – for which objective liability is provided – from other artificial intelligence for which the responsibility is culpable. With the proposal for a Regulation of 21st April 2021, governance in a political and institutional sense is entrusted with shaping the design of digital artifacts by conditioning ex ante their social impact and that of the devices in which they are incorporated.