The Impact of Artificial Intelligence Systems in Criminal Law: An Integral and Integrated Development Perspective

Introduction

Artificial intelligence (AI) tools are today a pervasive constant in everyone's working and social life. Consider, for istance, the “glocal” phenomenon of the so-called smart cities (or also digital, wired, virtual, intelligent, sustainable city), characterised by the use of information and communications technologies (ICT) to support the development of urban lifestyle (Dameri et al., 2019), or the processes of digitalisation of corporate governance and the first cases, already theorised, of total or partial replacement of company directors with intelligent machines (Scarchillo, 2019). The impetuous development of the new technologies consequently overwhelmed the various branches of law, both in terms of drafting new and detailed regulations on their use and in terms of adapting their own cardinal principles. Actually, in the field of administrative law, for example, it appeared necessary to regulate the digitisation strategies of public administrations (e-government) towards the «objectives of efficiency, effectiveness, economy, impartiality, transparency, simplification and participation in compliance with the principles of equality and non-discrimination, as well as for the effective recognition of the rights of citizens and businesses», as already established by the italian Digital Administration Code (L.d. no. 85/2005) in art. 12 c.1 (Faini, 2019, tba); and, moreover, a further example, this time in the field of civil law, is given by the European Parliament Resolution (February 16, 2017), entitled Resolution of Civil Law Rules on Robotics, where it is written for the first time the concept of electronic person to be attributed to artificial intelligence systems, at least in their most autonomous and sophisticated forms. But it is especially in the penal disciplines that the entry of AI will entail complex issues with not unequivocal solutions concerning (a) the normative and epistemological frictions arising from the attribution to such devices of the concepts of suitas, imputability and, more generally, criminal liability; (b) the ontological coherence between the actions carried out by these tools and certain cardinal principles of the legal systems of democratic countries, such as those of self-determination, materiality, offensiveness, the obligation of motivation for judicial measures, and the judge's subjection only to the law; (c) the perplexities raised by some doctrine (Pangallo, 2013; Cappellini, 2019), inherent to the functions performed by the sanction (re-educative, special- and general-preventive), since its recipients are machines and not humans. However, the impact that artificial intelligence has (and will have) in criminal law disciplines, is clearly shown in relation to the fields in which it will operate and the functions it will perform, with particular focus on crime prediction and repression, as well as changing the related codes, both substantive and procedural, towards a digital transition. Today's policy maker is therefore called to regulate this economic-social context, which is constantly and rapidly changing and in which AI, from the role of assistant to the human being, is increasingly demanding to become the undisputed dominus. In this transition, the most arduous task is undoubtedly that of balancing the delicate new-trade-offs between development and technological innovation with the inviolable rights of man, the principles of the legal system, the ethicality of bio-tech behaviour and, more generally, with everything that threatens non-anthropocentric social development.