| Abstract [eng] |
The integration of artificial intelligence (AI) technologies into diverse sectors, including medicine, engineering, and social media marketing, is significantly transforming daily consumer experiences. For instance, retail and consumer goods companies plan to allocate an average of 3.32% of their revenues to AI implementation starting in 2025. These investments are expected to target areas such as customer service, supply chain management, talent acquisition, and marketing innovations, emphasizing the expansion of AI beyond traditional IT applications (IBM, 2025a). While the use of AI in marketing is often seen as a way to reduce costs and improve efficiency (Ivanov, 2023), significant challenges related to consumer reactions to AI-generated content have also emerged. For example, studies indicate that consumers often struggle to distinguish between AI-generated and human-created advertisements, especially when using advanced systems like ChatGPT or Midjourney. AI-generated content frequently reaches a high level of quality, potentially triggering a sense of "technological eeriness" among users (Gu et al., 2024). Research also reveals that consumers often require additional context to identify the true author of the content. Nonetheless, AI-generated content is frequently perceived as more objective and less biased than human-created content, which can enhance the effectiveness of social media communications in certain contexts (Xie et al., 2024). Social media platforms are evolving into spaces not only for communication but also for AI-driven content creation, personalized recommendations, and real-time user behavior analysis (Mohamed et al., 2024). However, the use of these technologies remains controversial, particularly when users lack a clear understanding of how AI-powered systems make decisions (Scharowski et al., 2023). The primary objective of this master’s thesis is to compare the impact of AI-generated and human-created social media content on users‘ trust and intentions. To achieve this, the research is structured around five key goals: establish the relevance and significance of studying the impact of AI-generated and human-generated social media content on users’ trust and intentions; develop a conceptual model based on a theoretical analysis of the effects of AI-generated and human-generated social media content on users’ trust and intentions; design a methodology focused on revealing the impact of AI and human-generated social media content on users’ trust and intentions within social media; conduct an empirical study to interpret the findings within a scientific discussion, highlighting the limitations of the research and directions for future studies; provide practical recommendations for organizations and individuals seeking to effectively utilize AI and human-generated social media content. The empirical research revealed that perceived content vividness and persuasiveness have a positive impact on trust, while synthetic nature significantly negatively affects trust. It was also found that disclosing the authorship of AI-generated content negatively impacts trust, while revealing the authorship of human-created content has a positive effect. Contrary to expectations, perceived eeriness had a positive impact on trust, suggesting that respondents may associate unusual content with innovation and novelty. Additionally, the study found that the perception of AI as a co-author moderates the impact of author disclosure on trust, while the perception of AI as merely a tool does not significantly influence this relationship. The data analysis also indicated that changes in consumer trust positively influence their inclination to interact with the content, though this relationship was relatively weak, suggesting that trust is not the only factor driving consumer intentions. Moreover, regardless of subjective knowledge about AI tools or perceived ability to identify AI-generated content, only 32.5% of participants accurately identified the content author. Surprisingly, before the author was disclosed, respondents rated AI-generated content as more authentic, higher quality, and more trustworthy than human-created content, contrary to initial expectations. |
