Abstract

The European Union’s proposed AI Liability Directive is an ambitious attempt to address the unique challenges of AI, aiming to balance innovation with regulatory consistency and closely linked to definitions and risk categorizations in the AI Act. However, the minimal harmonization approach introduces procedural hurdles and raises doubts about its practical effectiveness. The Directive’s reliance on national implementation allows Member States flexibility, but this may result in legal fragmentation and differing liability standards. The present essay seeks to address the Directive’s key aspects, offering an analysis based on German and Italian perspectives. Through this comparative assessment, it presents critical observations to guide the continued development of the Directive in the aftermath of the AI Act