DEEPFAKES AND DIGITAL TRUST IN KOSOVO: DETERMINANTS OF PERCEIVED CREDIBILITY OF AI-GENERATED NEWS AND PUBLIC INTEREST CONTENT

This study examines the factors that shape how internet users in Kosovo judge the credibility of AI-generated news and public-interest content. A cross-sectional survey conducted in Albanian measured perceived credibility, source credibility, authenticity cues, and digital verification literacy using a five-point scale. The analysis conducted in IBM SPSS shows a clear pattern in how respondents rate content as more credible when they trust the source and when authenticity indicators are stronger, while stronger verification habits are associated with more cautious judgments. In the regression model, source credibility and authenticity cues were positive predictors of perceived credibility, whereas verification literacy was inversely related to perceived credibility, suggesting careful judgment. Overall, credibility online is shaped by what surrounds the content (source signals and authenticity cues) and what the user does (verification). The findings highlight practical steps for Kosovo to improve source transparency, strengthen authenticity cues, and promote simple verification habits for users.