Algorithmic Disinformation: Evaluating the Discourse of Generative AI GROK in the Context of Epistemic Uncertainty
ISTANBUL UNIVERSITESI SOSYOLOJI DERGISI-ISTANBUL UNIVERSITY JOURNAL OF SOCIOLOGY, cilt.46, sa.1, ss.187-218, 2026 (ESCI)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 46 Sayı: 1
- Basım Tarihi: 2026
- Doi Numarası: 10.26650/sj.2026.46.1.1147
- Dergi Adı: ISTANBUL UNIVERSITESI SOSYOLOJI DERGISI-ISTANBUL UNIVERSITY JOURNAL OF SOCIOLOGY
- Derginin Tarandığı İndeksler: Emerging Sources Citation Index (ESCI)
- Sayfa Sayıları: ss.187-218
- Anadolu Üniversitesi Adresli: Evet
Özet
This study examines the role of generative artificial intelligence within the contemporary disinformation ecosystem. Specifically, it analyzes how Grok constructs a discourse around verified disinformation contents on the social media platform X. The dataset consists of 43 posts shared bythe officialX account of the Turkish Presidency of Communications' Center for Combating Disinformation between August 1, 2025, and December 31, 2025. These posts, which collectively received more than 5.5 million views, identified 65 disinformation posts originating from various X accounts. The identified disinformation posts accumulated a total of 25,532,770 views and generated 5,924 user comments. Within these comments, 61 questions directed to Grok regarding the accuracy of the relevant posts were detected, along with Grok's 61 corresponding responses. These responses constitute the sample of the study. The analysis employs inductive coding within the framework of discourse analysis. Based on similarities and differences across Grok's responses, a simplified codingscheme was applied to evaluate: (1) whetherthe discourse functioned as corrective, ambiguous, or disinformation-reinforcing; (2) whether the epistemic tone signaled certainty, caution, or ambiguity; and (3) whether the responses relied on formal or informal sources of information. The findings, interpreted within the conceptual framework of disinformation and epistemic uncertainty, indicate that generative artificial intelligence does not operate as a neutral or purely corrective informational actor in the fight against disinformation. Rather, it functions as an intermediary communicative actor that can reconstruct ambiguity and, at times, reinforce disinformation, thereby increasing informational uncertainty within the communication ecosystem.