AI in digital news media : the impact of AI authorship and transparency on trust in news content and brands

Westerheide, Philine

http://hdl.handle.net/10400.14/55536

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
203971884.pdf		2.61 MB	Adobe PDF	Download

Send Feedback

Authors

Westerheide, Philine

Advisor(s)

Lancastre, Filipa

Abstract(s)

In response to growing economic pressures, news organizations have increasingly integrated Artificial Intelligence (AI) into the news production process to increase efficiency. Particularly in content creation, the core of journalism, this development raises the question of how readers perceive AI-generated content and whether it affects their trust in digital news amid rising media skepticism. Despite the increasing relevance, there is still no common academic understanding of how audiences interpret AI authorship. As a result, news organizations lack clear, evidence-based advice on how to effectively disclose AI involvement. By combining insights from technology acceptance, media research and trust theory, this study investigates how AI authorship disclosure and transparency statements influence readers' trust in news articles and brands in the context of digital journalism. The research is based on a quantitative experimental study with a 2 × 2 between-subjects design. The results show that hybrid authorship significantly increases trust in the news brand compared to AI-only authorship. No significant difference was found for news credibility. Transparency statements directly improved both brand trust and news credibility, but perceived transparency did not moderate the relationship between authorship type and news credibility as expected. Overall, the findings reveal the importance of demonstrating human involvement in AI- augmented journalism to maintain readers' trust. By analyzing readers' perceptions of hybrid journalism, an increasingly relevant practice, this study addresses a research gap and provides guidance for news organizations on how AI disclosure and transparency strategies can influence audience trust.

Face às crescentes pressões económicas, as organizações noticiosas estão a integrar cada vez mais a Inteligência Artificial (IA) na produção de notícias para aumentar a eficiência. Em especial na criação de conteúdos, o núcleo do jornalismo, esta evolução levanta questões sobre como os leitores percebem textos gerados por IA e se isso afeta a confiança nas notícias digitais. Apesar da sua importância, ainda não existe consenso académico nem orientações claras sobre como comunicar o uso da IA. Combinando teorias da aceitação da tecnologia, da confiança nos media e da teoria da confiança, esta dissertação analisa como a divulgação da autoria por IA e as declarações de transparência influenciam a confiança dos leitores em artigos e marcas de notícias no jornalismo digital. A investigação baseia-se num estudo experimental quantitativo com um desenho fatorial 2 × 2 entre sujeitos. Os resultados mostram que a criação híbrida aumenta significativamente a confiança na marca em comparação com a criação apenas baseada em IA. Não foram encontradas diferenças significativas na credibilidade percebida das notícias. As declarações de transparência aumentaram diretamente a confiança na marca e a credibilidade percebida, mas não moderaram a relação entre autoria e credibilidade. Os resultados revelam a importância de demonstrar o envolvimento humano no jornalismo com IA para manter a confiança dos leitores. Ao analisar perceções dos leitores sobre jornalismo híbrido, uma prática cada vez mais comum, este estudo preenche uma lacuna na investigação e oferece orientações sobre como estratégias de divulgação e transparência da IA podem influenciar a confiança.

Keywords

AI transparency Artificial intelligence Autoria híbrida Confiança nas notícias Confiança nos media Digital journalism Hybrid authorship Inteligência artificial Jornalismo digital Media trust Transparência da IA Trust in news