Uporaba orodij umetne inteligence pri odločanju v kazenskem pravu

magistrsko diplomsko delo

Andreja Gvozdić (Author), Miha Hafner (Mentor)

Abstract

Ena izmed bistvenih prvin, ki se jih zasleduje pri sodnem odločanju, je nepristransko sojenje. Vendar na ljudi neizogibno vplivajo marsikateri subjektivni elementi, poleg tega pa za lažje sprejemanje odločitev uporabljamo tudi kognitivne bližnjice. Posledično se zastavlja vprašanje, ali bi lahko z asistenco ali celo s prepustitvijo odločanja računalnikom, opremljenim z umetno inteligenco, dosegli objektivnejše in pravičnejše rezultate ter izboljšali kvaliteto sodnih odločb. Magistrsko diplomsko delo je vsebinsko razdeljeno na sklope z namenom prikazati vpliv umetne inteligence na odločanje v kazenskem postopku, tako da so najprej predstavljene pogoste napake sodniškega odločanja in vpliv subjektivnosti nanj, nato pa še poglavitne težave, ki se pojavljajo pri uporabi umetne inteligence. Sledi vpogled v ameriški pravni sistem, kjer se uporabe umetne inteligence že poslužujejo, s preverjanjem skladnosti tovrstnih algoritmov z načeli in pravicami, ki se zasledujejo v Evropski uniji, ter zaključi s poglavjem o transparentni umetni inteligenci.

Keywords

umetna inteligenca;nepristransko sojenje;pošteno sojenje;napake pri odločanju;odločanje z umetno inteligenco;težave umetne inteligence;COMPAS;skladnost umetne inteligence z evropskim pravom;transparentna umetna inteligenca;kazensko pravo;

Data

Language:	Slovenian
Year of publishing:	2025
Typology:	2.09 - Master's Thesis
Organization:	UL PF - Faculty of Law
Publisher:	[A. Gvozdić]
UDC:	343:004.8(043.2)
COBISS:	233022211
Views:	114
Downloads:	37
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	ǂThe ǂuse of artificial intelligence tools in criminal law decision-making
Secondary abstract:	One of the essential elements pursued in judicial decision-making is an impartial trial. However, people are inevitably influenced by many subjective elements, and cognitive shortcuts are also used to facilitate decision-making. As a result, the question arises as to whether assisting or even leaving decision-making to computers equipped with artificial intelligence could lead to more objective and fairer results and improve the quality of judicial decisions. The master's thesis is divided into sections in order to illustrate the impact of artificial intelligence on decision-making in criminal proceedings. First by presenting common errors in judicial decision-making and the impact of subjectivity on it, and then by presenting the main problems that arise from the use of artificial intelligence. This is followed by a look at the US legal system, where AI is already being used and an examination of the compatibility of such algorithms with the principles and rights pursued in the European Union. The thesis concludes with a chapter on transparent AI.
Secondary keywords:	Univerzitetna in visokošolska dela;
Type (COBISS):	Master's thesis/paper
Study programme:	0
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. v Ljubljani, Pravna fak.
Pages:	57 str.
ID:	26155474