o:10304 Human rights and civil law risks of artificial intelligence: a systematic analysis of surveillance, algorithmic bias, automated decision-making, and restrictions on freedom of expression Rizici veštačke inteligencije po ljudska prava i građanskopravnu zaštitu: sistematska analiza nadzora, algoritamske pristrasnosti, automatizovanog donošenja odluka i ograničavanja slobode izražavanja en Abstract: The development of artificial intelligence significantly shapes contemporary social processes, but at the same time opens up complex issues regarding the protection of fundamental human rights and the effectiveness of existing legal protection mechanisms. This paper provides a systematic analysis of the key risks that AI technologies produce in the areas of rights to privacy, equality, fair trial and freedom of expression, with particular reference to the role of civil protection as a complement to public law and regulatory instruments. The author analyzes the phenomenon of digital surveillance and the erosion of privacy due to the collection, analysis and predictive modeling of large data sets, including biometric information and data from daily user interactions, pointing out the limitations of the concept of informed consent and the need for individual private legal means of protection. Next, algorithmic bias is discussed as a mechanism that reproduces and deepens existing social inequalities, which threatens the right to equal treatment and raises the issue of civil liability for discriminatory outcomes of automated decision-making. Special attention is devoted to the application of AI in the judiciary, where non-transparent decision-making models, limited explainability and the risk of "automated suggestions" call into question the realization of procedural guarantees of a fair trial, as well as the possibility of compensation for damages due to the violation of procedural rights. The final part discusses the risks to freedom of expression, including the consequences of automated content moderation, personalized ranking of information and the generation of synthetic content, while pointing out the importance of civil law requirements for the protection of individual rights, reputation and dignity. By analyzing the relevant literature and the current regulatory framework, the paper identifies normative gaps and emphasizes the need for a coherent legal system that, in addition to public law supervision, also provides effective private law protection mechanisms. The paper concludes that sustainable regulation of artificial intelligence is possible only through a combination of technical transparency, democratized supervision and strengthening of procedural and material guarantees, including a developed system of civil liability in the digital environment. Apstrakt: Razvoj veštačke inteligencije značajno oblikuje savremene društvene procese, ali istovremeno otvara složena pitanja u pogledu zaštite temeljnih ljudskih prava i efektivnosti postojećih mehanizama pravne zaštite. Ovaj rad pruža sistematsku analizu ključnih rizika koje AI tehnologije proizvode u oblastima prava na privatnost, jednakost, pravično suđenje i slobodu izražavanja, sa posebnim osvrtom na ulogu građanskopravne zaštite kao dopune javnopravnim i regulatornim instrumentima. Autorka analizira fenomen digitalnog nadzora i erozije privatnosti usled prikupljanja, analiziranja i prediktivnog modelovanja velikih skupova podataka, uključujući biometrijske informacije i podatke iz svakodnevnih interakcija korisnika, ukazujući na ograničenja koncepta informisanog pristanka i potrebu za individualnim privatnopravnim sredstvima zaštite. Zatim se razmatra algoritamska pristrasnost kao mehanizam koji reprodukuje i produbljuje postojeće društvene nejednakosti, čime se ugrožava pravo na jednako postupanje i otvara pitanje građanskopravne odgovornosti za diskriminatorne ishode automatizovanog odlučivanja. Posebna pažnja posvećena je primeni AI u pravosuđu, gde netransparentni modeli odlučivanja, ograničena objašnjivost i rizik od „automatizovane sugestije“ dovode u pitanje ostvarivanje procesnih garancija pravičnog suđenja, ali i mogućnost naknade štete usled povrede procesnih prava. U završnom delu razmatraju se rizici po slobodu izražavanja, uključujući posledice automatizovanog moderiranja sadržaja, personalizovanog rangiranja informacija i generisanja sintetičkih sadržaja, pri čemu se ukazuje na značaj građanskopravnih zahteva za zaštitu prava ličnosti, ugleda i dostojanstva pojedinca. Analizom relevantne literature i važećeg regulatornog okvira rad identifikuje normativne praznine i naglašava potrebu za koherentnim pravnim sistemom koji, pored javnopravnog nadzora, obezbeđuje i delotvorne privatnopravne mehanizme zaštite. Rad zaključuje da je održiva regulacija veštačke inteligencije moguća samo kroz kombinaciju tehničke transparentnosti, demokratizovanog nadzora i jačanja procesnih i materijalnih garancija, uključujući razvijen sistem građanskopravne odgovornosti u digitalnom okruženju. Keywords: artificial intelligence, human rights, civil protection, digital surveillance, fair trial rights, freedom of expression, data protection, discrimination. Ključne reči: veštačka inteligencija, ljudska prava, građanskopravna zaštita, digitalni nadzor, pravo na pravično suđenje, sloboda izražavanja, zaštita podataka, diskriminacija. 1552101 1451-3870 1552099 10.5937/univmis2525038m 2026-09-16T11:45:02.610Z 44 yes 46 Jelena Matijašević University Business Academy in Novi Sad, Serbia 14634599 0000-0001-8068-0816 Maida Bećirović Alić University of Novi Pazar Novi Pazar, Serbia person 0000-0002-9738-6581 Jasmina Nikšić University of Novi Pazar Novi Pazar, Serbia person 0000-0002-8190-7565 Irma Mašović Muratović University of Novi Pazar Novi Pazar, Serbia person 0000-0003-0175-3584 application/pdf 464968 https://unilib.phaidrabg.rs/o:10304 no yes 16 70 92000001 211A01 Univerzitetska misao - casopis za nauku, kulturu i umjetnost, Novi Pazar 25 25 38 56 2026