20.500.12556/RUL-187886 Možna rešitev problematike detekcije besedil generiranih z UI na univerzah: uvedba sistema za sledljivost pozivov (promptov) in transparentno dokazovanje avtorstva magistrsko delo A potential solution to the AI-generated text detection challenge at universities: implementing a system for prompt traceability and transparent proof of authorship Magistrsko delo obravnava problematiko detekcije besedil, generiranih z umetno inteligenco, na univerzah. Osredotoča se na reševanje te problematike s predlaganim vmesnikom, ki omogoča transparenten vpogled v proces nastajanja dela z uporabo velikih jezikovnih modelov (VJM). Empirični del je bil izveden v treh fazah: v prvi fazi sta bila izvedena pregled literature o obstoječih metodah detekcije ter analiza Akta EU o umetni inteligenci in smernic Univerze v Ljubljani glede uporabe detektorjev. V drugi fazi so bile preučene rešitve tujih univerz na tem področju, v zadnji fazi pa sta bila izdelana teoretični načrt in prototip vmesnika za sledljivost pozivov in transparentno dokazovanje avtorstva. Izvedena je bila tudi primerjalna analiza za ugotovitev, ali takšna sistemska rešitev nudi boljšo alternativo za transparentno uporabo VJM-ov kot obstoječe metode detekcije. Raziskava potrjuje nezanesljivost klasičnih metod detekcije, saj te ne upoštevajo konteksta uporabe, izkazujejo sistemsko pristranskost ter ob preprostem preurejanju ali čiščenju besedil ne dajejo zanesljivih rezultatov. Kot učinkovita možna rešitev se je izkazal predlagani vmesnik, ki z vpogledom v celotno interakcijo natančno razlikuje med vsebinskim generiranjem in zgolj jezikovnim urejanjem besedila. S tem študente zanesljivo ščiti pred neupravičenimi obtožbami, pedagogom pa ponuja objektivno podlago za pravično vrednotenje študentskih del. The master's thesis addresses the problem of detecting AI-generated text in universities. It focuses on solving this problem with a proposed interface that enables transparent insight into the process of creating texts using large language models (LLMs). The empirical part was carried out in three phases: in the first phase, a literature review of existing detection methods and an analysis of the EU AI Act and the guidelines of the University of Ljubljana regarding the use of detectors were conducted. In the second phase, the solutions of foreign universities in this field were examined, and in the final phase, a theoretical plan and a prototype interface for prompt traceability and transparent proof of authorship were created. A comparative analysis was also conducted to determine whether such a systemic solution offers a better alternative for the transparent use of LLMs than existing detection methods. The research confirms the unreliability of classic detection methods, as they do not take into account the context of use, show systemic bias, and do not yield reliable results with simple editing or cleaning of texts. The proposed interface proved to be an effective possible solution, which, through insight into the entire interaction, accurately distinguishes between content generation and mere linguistic editing of the text. This reliably protects students from unjustified accusations, while offering educators an objective basis for fair evaluation of student work. detekcija besedil veliki jezikovni modeli sledljivost pozivov (promptov) dokazovanje avtorstva akademska integriteta text detection large language models prompt traceability proof of authorship academic integrity true true false T. Kopčavar Slovenski jezik Angleški jezik Magistrsko delo/naloga 2026-09-16 07:30:11 2026-09-16 07:30:16 2026-09-18 13:26:33 0000-00-00 00:00:00 2026 0 Ljubljana 0 Ljubljana 1 spletni vir (1 datoteka PDF (96 str.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.89:378(043.2) 208139 291629059 30845.pdf 30845.pdf 1 57E42FEFD1DA59402EEA804142FB586A 1e52f4331e49b69c41cfa2dd6445b15c4aa8a0bbc1e022e6cc10990d03ea82d5 aa2a910d-b18f-11f1-8bc5-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=247710 Fakulteta za družbene vede 0 0 0