20.500.12556/DKUM-99107 "Komu zaupamo svoje skrbi?": umetna inteligenca kot vir čustvene podpore magistrsko delo "In Whom Do We Confide?": Artificial Intelligence as a Source of Emotional Support Generativna umetna inteligenca vse pogosteje prevzema vlogo sogovornika pri čustvenih temah, pri čemer ostaja nejasno, kakšne psihološke posledice ima takšna raba. Namen raziskave je bil preučiti povezave med uporabo generativne umetne inteligence za čustvene teme, zaupanjem vanjo, zaznano čustveno podporo, zaznanim stresom in psihološko stisko ter preveriti, ali zaznana čustvena podpora po vzoru modela blažilnega učinka ublaži odnos med stresom in stisko. V spletni raziskavi je sodelovalo 198 polnoletnih udeležencev (M = 33,27; SD = 14,62), od katerih jih je 174 poročalo o uporabi klepetalnikov z umetno inteligenco, 108 pa o njihovi uporabi za čustvene teme. Podatke smo analizirali z linearnimi regresijami ter analizama moderacije in mediacije, odgovore 88 udeležencev na odprta vprašanja pa s tematsko analizo. Uporaba za čustvene teme in zaupanje sta pozitivno napovedovala zaznano čustveno podporo, zaznani stres pa psihološko stisko. Zaznana čustvena podpora ni bila povezana z nižjo psihološko stisko, prav tako nismo našli statistično značilnih dokazov, da bi moderirala odnos med stresom in stisko ali posredovala v odnosu med uporabo in stisko. Pogostejša uporaba za čustvene teme je bila povezana z višjo psihološko stisko. Tematska analiza je pokazala, da udeleženci klepetalnike doživljajo predvsem kot sredstvo za trenutno čustveno razbremenitev in ubeseditev težav, ne pa kot vir zanesljivo razpoložljive podpore. Generative artificial intelligence is increasingly taking on the role of a conversational partner in discussions of emotional topics, while the psychological consequences of such use remain unclear. The aim of this study was to examine the relationships between the use of generative artificial intelligence for emotional topics, trust in it, perceived emotional support, perceived stress, and psychological distress, and to test whether perceived emotional support buffers the relationship between stress and distress in line with the stress-buffering model. The online study included 198 adult participants (M = 33.27; SD = 14.62), of whom 174 reported using AI chatbots and 108 reported using them for emotional topics. The data were analysed using linear regression, moderation, and mediation analyses, while the responses of 88 participants to open-ended questions were examined using thematic analysis. Use for emotional topics and trust positively predicted perceived emotional support, while perceived stress positively predicted psychological distress. Perceived emotional support was not associated with lower psychological distress, and no statistically significant evidence was found that it moderated the relationship between stress and distress or mediated the relationship between use and distress. More frequent use for emotional topics was associated with higher psychological distress. The thematic analysis showed that participants experienced chatbots primarily as a way to obtain momentary emotional relief and as help with articulating their difficulties, rather than as a source of reliably available support. umetna inteligenca zaznana čustvena podpora stres psihološka stiska blažilni učinek socialne podpore artificial intelligence perceived emotional support stress psychological distress stress-buffering effect of social support true true false N. Ritonja Slovenski jezik Angleški jezik Magistrsko delo/naloga 2026-07-31 14:40:31 2026-09-04 09:59:18 2026-09-05 03:09:29 0000-00-00 00:00:00 2026 0 Maribor 0 Maribor 1 spletni vir (1 datoteka PDF (XII, 97 str.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 159.942:004.8(043.2) 290012419 MAG_Ritonja_Nika_2026.pdf MAG_Ritonja_Nika_2026.pdf 1 411E23655E157F9372464052D8504714 6eeedd2a6a0ef0a1c72e7185f186a5043cf6902c57d0f2354023e080983c734c 3f3136cc-8ce2-11f1-892d-00155d0001be https://dk.um.si/Dokument.php?lang=slv&id=208815 Filozofska fakulteta 0 0 0