Review of the book 'Three liability regimes for Artificial Intelligence: algorithmic actants, hybrids, crowds', by Anna Beckers and Gunther Teubner (Oxford: Hart, 2021)

Abstract

The book Three Liability Regimes for Artificial Intelligence: Algorithmic Actants, Hybrids, Crowds, by Anna Beckers and Gunther Teubner, published in 2021 by Hart Publishing, proposes a dogmatic reconfiguration of tort law in light of the risks inherent to artificial intelligence systems. The work addresses the extensive responsibility gap generated by AI, which manifests itself in the unpredictability of algorithmic actions and the inadequacy of traditional legal categories. The authors reject both extreme solutions — either a universal one-size-fits-all approach to liability or excessive sectoral fragmentation — and instead advocate a model that calibrates legal responsibility according to a typology of fundamental digital risks. The methodology combines insights from information technology studies, comparative legal sociology, and risk analysis, introducing the concept of “socio-digital institutions” as intermediate disciplines that mediate between technology and the legal order.