Grok, is this true? w/ Akin Unver

Grok, is this true? w/ Akin Unver

by Akin Unver

Humanities, Social Sciences & Thought Media & Information JBCTMedia studiesJBCT5Disinformation and misinformation
🎙 Akin Unver 👥 6K 📅 May 21, 2026 ⏱ 42 min 👁 169 📄 expert opinion 🧭 2026-08-16
Available in: English (current) Français

Keywords

Grokperformative fact-checkingTurkish information spaceAI-assisted verificationpolarization

Summary

Akin Unver, a political scientist, presents his ongoing research on the use of Grok, an AI chatbot embedded in X (Twitter), for fact-checking in the Turkish information space. He introduces the concept of ‘performative fact-checking,’ where users invoke Grok not to seek truth but to win arguments and demonstrate their correctness to an audience. The study collected around 100,000 organic posts invoking Grok over a year, using regex and keyword classifiers to distinguish sincere from performative invocations. Preliminary findings suggest that performative fact-checks lead to higher replies, reposts, and views but lower likes, indicating that they fuel controversy and polarization rather than fostering open-mindedness. The research challenges the assumption that AI-assisted fact-checking promotes democratic values and civility. Turkey is chosen as a case study due to its polarized, crisis-prone, and multilingual information environment, with high X usage and low trust in state information. The project is part of a larger EU-funded consortium studying foreign information manipulation and interference. Unver discusses the theoretical foundations, data collection challenges, and implications for policy, suggesting that AI tools like Grok may reinforce existing beliefs rather than encourage open-mindedness.

184 words

Critical Evaluation

Value of the Information & Strength of the Argument

The lecture provides valuable insights into the emerging phenomenon of AI-assisted fact-checking in social media, introducing the novel concept of ‘performative fact-checking’ and supporting it with preliminary empirical evidence. The argumentation is coherent, linking theoretical frameworks from political science and communication studies to concrete observations. However, the findings are preliminary and based on a single-country case, limiting generalizability. The speaker acknowledges the ongoing nature of the research and the need for comparative data. The discussion of data collection methods and challenges adds credibility, but the lack of detailed statistical analysis or peer review tempers the strength of the claims.

Scientific Rigor, Source Quality, Title Accuracy

The lecture demonstrates scientific rigor through a clear research design, transparent data collection methods, and reference to established literature on misinformation and performativity. The speaker cites relevant academic works (e.g., Bonzel et al., Holes et al., Simson et al.) and mentions collaborations with fact-checking organizations. However, specific sources are not provided in the description, limiting verifiability. The title accurately reflects the content, focusing on Grok and fact-checking dynamics. The presentation is part of a series on ‘Tech for Open Minds,’ which aligns with the theme of examining technology’s impact on open-mindedness. No comments were provided for analysis.

211 words

Title / Content Match

The title accurately reflects the content, focusing on the use of Grok for fact-checking in Turkish information space.

Quality & Reliability

7/10

The lecture presents preliminary findings from an ongoing research project, with clear methodology and data collection details, but lacks peer review and full transparency on data analysis.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

Concurring Sources

Contribution & Novelties

The lecture introduces the concept of ‘performative fact-checking’ as a distinct social practice, providing a new lens to understand AI-assisted verification in polarized online environments. It challenges the assumption that AI fact-checking tools inherently promote truth-seeking and civility, suggesting instead that they may be weaponized in partisan debates. The research offers empirical evidence from a unique case (Turkey) and proposes a framework for studying similar dynamics in other contexts.

Pour aller plus loin :

100 words

Radar Profile

The radar profile shows high scores in quantity of information and technical level, indicating a detailed and methodologically sound presentation. The lower score in quality of information reflects the preliminary nature of the findings and lack of peer review. Overall, the lecture is informative and technically competent, but the reliability is moderate due to ongoing research.

Reliability 7/10