Evaluating the Impact of AI-Generated News on Public Trust and Media Consumption

Thesis Abstract

The rapid integration of artificial intelligence (AI) technologies into news production has fundamentally transformed media landscapes, raising critical questions about the implications for public trust and media consumption patterns. This study addresses the burgeoning concern that AI-generated news, while enhancing efficiency and personalization, may adversely influence the credibility of news sources and alter audience engagement in complex ways. The primary aim is to evaluate the impact of AI-generated news on public trust and media consumption behaviors, with specific objectives including (1) examining the level of public trust in AI-generated versus human-curated news; (2) identifying factors that influence trust in AI-generated news content; (3) analyzing changes in media consumption patterns attributable to AI news; and (4) assessing the moderating roles of demographic variables and media literacy levels. Employing a mixed-methods research design, the quantitative phase involved a survey administered to a stratified random sample of 500 active social media users from urban areas with high digital media exposure, while qualitative insights were garnered through semi-structured interviews with 20 media professionals and technology experts. Data collection instruments included a validated questionnaire measuring trust, perceived credibility, and media consumption preferences, alongside interview guides developed through prior literature. Quantitative data were analyzed using multiple regression analysis to identify predictors of trust and media use, and ANOVA tests to examine differences across demographic groups. Thematic analysis was employed to interpret qualitative data, providing contextual understanding of perceptions surrounding AI-generated news. It is anticipated that findings will reveal a nuanced relationship where perceptions of AI-generated news influence trust levels variably across demographic segments, with younger audiences exhibiting greater openness yet heightened skepticism in certain contexts. The regression models are expected to show that perceived accuracy, transparency, and source attribution significantly predict public trust, while media literacy moderates these relationships. Additionally, the study expects to observe a shift toward personalized media consumption, with AI shaping preferences for tailored content, potentially impacting traditional media engagement. The insights gained will contribute to theoretical discourse by extending the Technology Acceptance Model (TAM) within the context of AI news, and will illuminate the interplay between trust, technology familiarity, and media behavior. This research offers significant contributions to knowledge by providing empirically grounded insights into trust dynamics in AI-mediated news environments, highlighting the socio-technical factors influencing media consumption. The findings can guide media organizations, policymakers, and technologists in designing transparent AI systems and fostering media literacy to mitigate trust erosion. It underscores the importance of establishing ethical standards and user-centric approaches in AI news deployment to sustain democratic information access. In conclusion, the study emphasizes that while AI-generated news has the potential to revolutionize media practices, careful consideration of trust-building mechanisms and audience education is essential. Recommendations include promoting transparency about AI content generation processes, developing regulatory frameworks for ethical AI use, and implementing media literacy programs to enhance user understanding of AI's role in news production. Future research should explore longitudinal effects of AI news adoption and investigate cross-cultural variations in trust perceptions, thereby advancing comprehensive knowledge in this evolving digital media landscape.