The study of civil liability within the algorithmic society and artificial intelligence requires a careful and profound reconsideration of classical legal principles. The new challenges posed by the algorithmic processing of data, by machinic subjectivation, and by the possible personhood of AI necessitate a regulatory update that takes into account the complexity of the phenomenon; while, on the one hand, technological innovation opens up new opportunities, on the other it generates risks that require adequate and timely responses. The widespread implementation of new technologies is leading to the dissolution of law as traditionally understood, directing the normative and regulatory system towards new trajectories that transcend the very concepts of norm and fact. Unlike in the past, neo-law is no longer pre-existing to the event and subsequently applied to the concrete case; rather, it is self-produced by the very reality it contributes to structuring.The emergence of autonomous artificial intelligences and software agents endowed with decision-making capabilities has triggered a profound transformation in the foundations of law, calling into question the traditional categories of legal subjectivity and normative responsibility. “Algorithmic normativity”, in which rules are not only imposed by a human legislator but are generated and applied by autonomous computational systems, introduces unprecedented questions that prompt inquiry into whether algorithms may be regarded as legal subjects and endowed with their own normative status. Furthermore, it raises the issue of how responsibility is to be distributed within automated decision-making ecosystems. In this context, concepts such as legal hybrids and algorithmic actants emerge entities operating in a grey area between technological instrumentality and operational autonomy, thereby challenging the dichotomy between legal subject and object. Through a new machinic language that simultaneously becomes normative code, computational logic and predictive models come to supersede legislative text. In the face of this metamorphosis which necessitates a rethinking of the sources of law, regulatory dynamics, and guarantees of transparency and accountability the jurist can no longer rely exclusively on consolidated traditional normative models, but must move with ease across law, technology, and philosophy, adopting an interdisciplinary approach capable of balancing technological innovation with the protection of fundamental rights. In an era in which the norm becomes code and code becomes norm, the challenge is not merely legal but epistemological, bearing upon the future of legal rationality in a world governed by machines. Civil liability arising from the use of algorithms and the damages caused by AI deserves careful examination with regard to the various types and hypotheses of algorithmic harm, including the manipulation of information. Consideration is given to the possibility of attributing legal responsibility for potential damages to developers, producers, or end-users; the introduction of a new regime of strict liability is envisaged, and some even venture the creation of a legal personality for AI. Such normative evolution materialises through regulatory autopoiesis within transnational socio-technical networks, giving rise to forms of private normativity.
Lo studio della responsabilità civile nella società algoritmica e dell’intelligenza artificiale richiede un attento e profondo ripensamento dei principi giuridici classici. Le nuove sfide poste dal trattamento algoritmico dei dati, dalla soggettivizzazione macchinica e dalla possibile personhood dell’AI obbligano a un aggiornamento normativo che tenga conto della complessità del fenomeno; se da un lato, infatti, l’innovazione tecnologica apre nuove opportunità, dall’altro genera rischi che necessitano di risposte adeguate e tempestive. L’implementazione massiva delle nuove tecnologie sta determinando la dissoluzione del diritto inteso in senso classico, dirigendo il sistema normativo e regolamentare verso nuovi approdi che superano il concetto di norma e di fatto. Il neo-diritto diversamente dal passato non è più preesistente al fatto ed applicato al caso concreto, ma, viene autoprodotto dalla stessa realtà che contribuisce a strutturare. L’emergere di intelligenze artificiali autonome e di agenti software dotati di capacità decisionale ha innescato una profonda trasformazione nelle fondamenta del diritto, mettendo in discussione le tradizionali categorie della soggettività giuridica e della responsabilità normativa. La “normatività algoritmica”, in cui le regole non sono solo imposte da un legislatore umano, ma generate e applicate da sistemi computazionali autonomi, introduce interrogativi inediti che portano a domandarsi se gli algoritmi possono essere considerati soggetti giuridici e detenere uno status normativo proprio. Inoltre, ci si chiede come si distribuisce la responsabilità in ecosistemi decisionali automatizzati. In tale contesto emergono concetti come ibridi giuridici e attanti algoritmici, entità che operano in una zona grigia tra strumentalità tecnologica e autonomia operativa, sfidando la dicotomia tra soggetto e oggetto giuridico. Attraverso un nuovo linguaggio macchinico che diventa allo stesso tempo anche codice normativo si supera con la logica computazionale e i modelli predittivi il testo legislativo. Innanzi a questa metamorfosi, che impone un ripensamento delle fonti del diritto, delle dinamiche regolatorie e delle garanzie di trasparenza e accountability, il giurista non può più ancorarsi esclusivamente ai consolidati modelli normativi tradizionali, ma deve muoversi con disinvoltura tra diritto, tecnologia e filosofia, adottando un approccio interdisciplinare in grado di bilanciare l’innovazione tecnologica con la tutela dei diritti fondamentali. In un’epoca in cui la norma diventa codice e il codice diventa norma, la sfida non è solo giuridica, ma epistemologica sulla quale sarà il futuro della razionalità giuridica in un mondo governato dalle macchine. La responsabilità civile derivante dall’uso degli algoritmi e i danni causati dall’AI merita di essere esplorata, rispetto alle varie tipologie ed ipotesi di danno algoritmico, inclusa la manipolazione dell’informazione. Si ragiona sull’eventualità di attribuire la responsabilità giuridica di eventuali danni agli sviluppatori, ai produttori, all’utente finale, si ipotizza la realizzazione di un nuovo regime di responsabilità oggettiva e qualcuno azzarda la creazione di una personalità giuridica per l’AI ; una evoluzione normativa che si sostanzia attrverso l’autopoiesi regolamentare all’interno di reti socio-tecniche transnazaionali, dando vita a forme di normatività privata.
La responsabilita’ civile da trattamento algoritmico dei dati, integrazione tecnologica ed evoluzione normativa. La sostenibilita’ delle reti socio - tecniche / Fontana, G.. - (2026 Jul 23).
La responsabilita’ civile da trattamento algoritmico dei dati, integrazione tecnologica ed evoluzione normativa. La sostenibilita’ delle reti socio - tecniche
Fontana, Gino
2026-07-23
Abstract
The study of civil liability within the algorithmic society and artificial intelligence requires a careful and profound reconsideration of classical legal principles. The new challenges posed by the algorithmic processing of data, by machinic subjectivation, and by the possible personhood of AI necessitate a regulatory update that takes into account the complexity of the phenomenon; while, on the one hand, technological innovation opens up new opportunities, on the other it generates risks that require adequate and timely responses. The widespread implementation of new technologies is leading to the dissolution of law as traditionally understood, directing the normative and regulatory system towards new trajectories that transcend the very concepts of norm and fact. Unlike in the past, neo-law is no longer pre-existing to the event and subsequently applied to the concrete case; rather, it is self-produced by the very reality it contributes to structuring.The emergence of autonomous artificial intelligences and software agents endowed with decision-making capabilities has triggered a profound transformation in the foundations of law, calling into question the traditional categories of legal subjectivity and normative responsibility. “Algorithmic normativity”, in which rules are not only imposed by a human legislator but are generated and applied by autonomous computational systems, introduces unprecedented questions that prompt inquiry into whether algorithms may be regarded as legal subjects and endowed with their own normative status. Furthermore, it raises the issue of how responsibility is to be distributed within automated decision-making ecosystems. In this context, concepts such as legal hybrids and algorithmic actants emerge entities operating in a grey area between technological instrumentality and operational autonomy, thereby challenging the dichotomy between legal subject and object. Through a new machinic language that simultaneously becomes normative code, computational logic and predictive models come to supersede legislative text. In the face of this metamorphosis which necessitates a rethinking of the sources of law, regulatory dynamics, and guarantees of transparency and accountability the jurist can no longer rely exclusively on consolidated traditional normative models, but must move with ease across law, technology, and philosophy, adopting an interdisciplinary approach capable of balancing technological innovation with the protection of fundamental rights. In an era in which the norm becomes code and code becomes norm, the challenge is not merely legal but epistemological, bearing upon the future of legal rationality in a world governed by machines. Civil liability arising from the use of algorithms and the damages caused by AI deserves careful examination with regard to the various types and hypotheses of algorithmic harm, including the manipulation of information. Consideration is given to the possibility of attributing legal responsibility for potential damages to developers, producers, or end-users; the introduction of a new regime of strict liability is envisaged, and some even venture the creation of a legal personality for AI. Such normative evolution materialises through regulatory autopoiesis within transnational socio-technical networks, giving rise to forms of private normativity.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
