20.500.12556/RUL-171879 Globoki ponaredki v novinarstvu: analiza uredniških politik v slovenskih uredništvih magistrsko delo Deepfakes in journalism: an analysis of editorial policies in Slovenian media Tehnologija globokih ponaredkov, ki temelji na generativni umetni inteligenci in omogoča ustvarjanje ter poustvarjanje prepričljivih avdiovizualnih podob, predstavlja enega največjih izzivov sodobnega novinarstva. Čeprav gre za razmeroma nov pojav, že pomembno vpliva na zaupanje javnosti v medije ter na širše razumevanje resnice v sodobnem digitalnem medijskem okolju. Magistrsko delo obravnava dvojno vlogo novinarstva v odnosu do omenjene tehnologije: novinarji nastopajo tako kot potencialni uporabniki orodij za ustvarjanje globokih ponaredkov, kot tudi varuhi resnice, ki morajo globoke ponaredke prepoznavati in izločati iz medijskega ekosistema. Osrednji del naloge zato predstavlja prenos tehnologije globokih ponaredkov v kontekst novinarstva tako z vidika uporabnosti tehnologije pri vsakdanjem novinarskem delu, kot tudi z vidika prepoznavanja sintetičnih avdiovizualnih podob. V sklepnem delu problematiko globokih ponaredkov vpeljem še v slovenski medijski prostor ter v okviru raziskave raziščem, kako se slednji prilagaja novi medijski realnosti, v kateri je meja med resničnim in ponarejenim precej zabrisana. Naloga tako ponuja aktualen vpogled v trenutno stanje, percepcijo grožnje ter odzive slovenskih uredništev na izzive, povezane z uporabo in zlorabo generativne umetne inteligence. The deepfake technology is based on generative artificial intelligence and enables the creation and manipulation of convincing audiovisual materials that pose one of the greatest challenges to contemporary journalism. Although being a relatively recent phenomenon, it already significantly affects public trust in the media and the broader understanding of truth within the modern digital media environment. This master's thesis examines the dual role of journalism in relation to this technology: journalists act both as potential users of tools for creating deepfakes and as guardians of truth, who must be able to identify and eliminate deepfakes from the media ecosystem. The core of the thesis focuses on transferring the concept of deepfake technology into the context of journalism, both from the perspective of its potential application in everyday journalistic work and from the perspective of recognizing synthetic audiovisual materials. In the final part, the issue of deepfakes is contextualized within the Slovenian media landscape. Through empirical research, the thesis explores how Slovenian media outlets are adapting to a new media reality in which the line between authentic and fabricated content is increasingly blurred. Thus, the thesis provides an up–to–date insight into the current state, the perception of the threat, and the responses of Slovenian editorial offices to the challenges associated with the use and misuse of generative artificial intelligence. globoki ponaredki generativna umetna inteligenca sintetične avdiovizualne podobe novinarstvo uredniške politike deepfakes generative artificial intelligence synthetic audiovisual media journalism editorial policies true true false P. Krištofelc Slovenski jezik Angleški jezik Magistrsko delo/naloga 2025-09-04 07:30:59 2025-09-04 07:31:02 2025-09-24 03:48:08 0000-00-00 00:00:00 2025 0 Ljubljana 0 Ljubljana 1 spletni vir (1 datoteka PDF (125 str.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 070:004.8(043.2) 179076 249747203 21480.pdf 21480.pdf 1 0F64D5AD60B17C108876D606F2E67C43 3ba199fe720a11191fa93c72328573d8409121eb53a4802525f212963f41d60c 3254a3a3-8950-11f0-9328-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=215444 Fakulteta za družbene vede 0 0 0