The topic of technological and digital evolution of public administration has for years been at the center of pronouncements and studies by jurisprudence and doctrine, which have been questioning the effect of information technology tools, especially those concerning the automation of public choices, on administrative procedure and public decisions. Indeed, in the lack of legal provisions, it has been on judges and scholars the identification and the adaptation of rules and principles provided for by the Administrative Proceeding Law. The object of this thesis, however, is different. In fact, it was decided to turn attention not so much to the simpler algorithms that so far appear to be used by public bodies for reasons of efficiency and celerity, but rather the use of so-called predictive technologies for cognitive or instructional purposes. Such technologies constitute a set of methods and algorithms that, partly based on artificial intelligence and self-learning methods, make use of historical data and statistics to extract “predictive value”. Just as in the private sector, where we see their application already in various fields such as finance, health care, logistics and cybersecurity, public agencies and government are also making use of them in various and disparate areas and, in particular, whenever it is called upon to make decisions that require it to have a certain predictive capacity. The present study, therefore, investigates the impact of such predictive technologies on public power and questions how the cognitive and evaluative activity of the public administration is changing, even when it does not result in an individual decision, but takes the form of an activity carried out as a preparatory public function or, more generally, in an autonomous cognitive activity. This is because, as emerged from the study of the application cases, public administrations seem to make use of predictive technologies to carry out control, supervisory and planning activities in order to intervene, in a precautionary and anticipatory key, to prevent the occurrence of a negative event. The basic question of the present research is, more generally, to ask whether it is possible to identify, a new power or function of the administration or whether, on the other hand, the use of such tools, however pervasive insofar as they are also based on the exploitation of citizens’ personal data, can nevertheless be traced back to known conceptual categories or functions already exercised by it. From the outcome of this question derive several consequences in terms of legal policy, relating in particular to the role of regulation about the use of predictive technologies in public decision-making processes. In fact, just as noted for automated decisions, once again the legislature, national and European, has remained essentially silent and the enucleation of principles or, better said, recommendations has been left essentially to soft law instruments of supranational character.
Il tema dell’evoluzione tecnologica e digitale della pubblica amministrazione è da anni al centro di pronunce e studi della giurisprudenza e della dottrina, che si interrogano sull’effetto degli strumenti informatici, specialmente quelli riguardanti l’automazione delle scelte pubbliche, sul procedimento e sul provvedimento amministrativo. Nel totale silenzio del legislatore, infatti, è stato compito dei giudici e degli studiosi tentare di ricostruire un “sistema” che prevedesse l’aggiornamento e l’adattamento di regole previste dalla legge sul procedimento e, più in generale, facenti parte dello statuto del diritto amministrativo. L’oggetto della presente tesi è tuttavia diverso. Si è deciso infatti di rivolgere l’attenzione non tanto agli algoritmi più semplici che fino ad ora risultano utilizzati all’interno del procedimento, per ragioni di efficienza e celerità, quanto piuttosto l’uso delle c.d. tecnologie predittive a fini conoscitivi o istruttori da parte dell’amministrazione. Tali tecnologie costituiscono un insieme di metodi e algoritmi che, anche sulla base di intelligenza artificiale e metodi di autoapprendimento, si avvalgono di dati storici e statistiche per estrarre da essi un “valore predittivo”. Così come nel settore privato, dove si assiste alla loro applicazione già in diversi campi quali quello della finanza, della sanità, della logistica e della sicurezza informatica, anche la pubblica amministrazione si avvale di esse in diversi e disparati settori e, in particolare, ogni qual volta sia chiamata ad assumere decisioni che le richiedono una certa capacità previsionale. Il presente studio, quindi, indaga l’impatto di tali tecnologie predittive sul potere pubblico e si interroga su come stia cambiando l’attività conoscitiva e valutativa della pubblica amministrazione, anche qualora essa non si traduca in un provvedimento, ma si concretizzi in un’attività svolta in funzione preparatoria di un procedimento amministrativo o, più in generale, in un’attività conoscitiva autonoma. Ciò in quanto, così come emerso dallo studio delle fattispecie applicative, la pubblica amministrazione sembra avvalersi delle tecnologie predittive per svolgere attività di controllo, di vigilanza e di programmazione al fine di intervenire, in chiave cautelativa e anticipatoria, per impedire la verificazione di un evento negativo. L’interrogativo di fondo della presente ricerca consiste, più in generale, nel chiedersi se sia possibile individuare, un nuovo potere o una nuova funzione dell’amministrazione o se invece l’uso di tali strumenti, per quanto pervasivo in quanto basato anche sullo sfruttamento dei dati personali dei cittadini, sia comunque riconducibile a categorie concettuali note o a funzioni già esercitate da questa. Dall’esito di questa domanda derivano diverse conseguenze in punto di politica del diritto, relative in particolare al ruolo della regolazione con riferimento all’uso delle tecnologie predittive nei processi decisionali pubblici. Difatti, così come rilevato per le decisioni automatizzate, ancora una volta il legislatore, nazionale ed europeo, è rimasto sostanzialmente silente e la enucleazione di principi o, per meglio dire, di raccomandazioni è stata rimessa sostanzialmente a strumenti di soft law dal carattere sovranazionale.
NUOVE TECNOLOGIE E ATTIVITÀ PREDITTIVA DELLA PUBBLICA AMMINISTRAZIONE
LEGGIO, MICHELA
2023
Abstract
The topic of technological and digital evolution of public administration has for years been at the center of pronouncements and studies by jurisprudence and doctrine, which have been questioning the effect of information technology tools, especially those concerning the automation of public choices, on administrative procedure and public decisions. Indeed, in the lack of legal provisions, it has been on judges and scholars the identification and the adaptation of rules and principles provided for by the Administrative Proceeding Law. The object of this thesis, however, is different. In fact, it was decided to turn attention not so much to the simpler algorithms that so far appear to be used by public bodies for reasons of efficiency and celerity, but rather the use of so-called predictive technologies for cognitive or instructional purposes. Such technologies constitute a set of methods and algorithms that, partly based on artificial intelligence and self-learning methods, make use of historical data and statistics to extract “predictive value”. Just as in the private sector, where we see their application already in various fields such as finance, health care, logistics and cybersecurity, public agencies and government are also making use of them in various and disparate areas and, in particular, whenever it is called upon to make decisions that require it to have a certain predictive capacity. The present study, therefore, investigates the impact of such predictive technologies on public power and questions how the cognitive and evaluative activity of the public administration is changing, even when it does not result in an individual decision, but takes the form of an activity carried out as a preparatory public function or, more generally, in an autonomous cognitive activity. This is because, as emerged from the study of the application cases, public administrations seem to make use of predictive technologies to carry out control, supervisory and planning activities in order to intervene, in a precautionary and anticipatory key, to prevent the occurrence of a negative event. The basic question of the present research is, more generally, to ask whether it is possible to identify, a new power or function of the administration or whether, on the other hand, the use of such tools, however pervasive insofar as they are also based on the exploitation of citizens’ personal data, can nevertheless be traced back to known conceptual categories or functions already exercised by it. From the outcome of this question derive several consequences in terms of legal policy, relating in particular to the role of regulation about the use of predictive technologies in public decision-making processes. In fact, just as noted for automated decisions, once again the legislature, national and European, has remained essentially silent and the enucleation of principles or, better said, recommendations has been left essentially to soft law instruments of supranational character.File | Dimensione | Formato | |
---|---|---|---|
phd_unimib_863380.pdf
accesso aperto
Dimensione
1.44 MB
Formato
Adobe PDF
|
1.44 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/75463
URN:NBN:IT:UNIMIB-75463