Masterarbeit im Studiengang Informatik von Nadja Bauer: Acceptance of AI-Based Hate Speech Moderation on Social Media: The Impact of Familiarity with AI Systems, IT Knowledge, and Social Media Use

As the prevalence of hate speech on social media continues to rise, there is an increasing rationale for the implementation of AI-based moderation systems. This study addresses the existing research gap in understanding how individual factors, including user familiarity with AI, IT proficiency, and social media use influence attitudes towards AI-based hate speech moderation. Based on an extensive literature review, the study develops three hypotheses: General familiarity with AI systems negatively correlates with acceptance of AI-based hate speech moderation (H1); higher levels of IT knowledge positively correlate with acceptance of AI-based hate speech moderation (H2); and a greater extent of social media use is positively associated with acceptance of AI-based moderation (H3). To explore these hypotheses, an online survey on the use of AI to address hate speech on social media is conducted (N = 115), and the data is analyzed using a multiple linear regression model. The results show that familiarity with AI systems does not significantly influence acceptance, while frequent social media use shows a slight, non-significant positive trend toward acceptance. In contrast, a notable negative correlation was found between IT knowledge and the acceptance of AI-based hate speech moderation. These findings provide both theoretical insights and practical implications for the use of AI-based content moderation systems on social media.

Betreuer: Julian Bäumler, M.A.
Prüfer: Prof. Dr. Dr. Christian Reuter