Too Good to Be True? Generalization Challenges in LLM-Based Fake News Detection

• Gustavo Gambarra UFCG
• Caio L. M. Jerônimo UEPB
• Claudio E. C. Campelo UFCG

DOI: https://doi.org/10.5753/webmedia.2025.15163

Resumo

Although recent studies report high accuracy in fake news classification using supervised models, such models often rely heavily on spurious correlations in training datasets and fail to generalize to unseen contexts. A smaller body of work has explored more realistic approaches based on linguistic cues, such as subjectivity, aiming to reduce the reliance on dataset-specific content learned during training. In this study, we assess the capabilities of four state-of-the-art large language models (LLMs) — Llama3.1-70B, Claude 3.5 Sonnet, GPT-4o, and Mistral Large 2 — in identifying fake news based solely on linguistic features, without access to external verification. Two experimentswere conducted: one with generic prompts and another with fine-grained instructions and examples, including classification of false excerpts into categories like untrue facts, exaggerations, and incorrect named entities. Despite the well-known linguistic capabilities of modern LLMs, our results show consistently low performance across both experiments. This outcome supports the hypothesis that detecting disinformation based solely on content or linguistic cues, without external factual grounding, is far more challenging than commonly reported, and that current evaluation protocols may overestimate the generalization capabilities of fake news classifiers.

Palavras-chave: fake news, large language models, fake news classification

Referências

Hunt Allcott and Matthew Gentzkow. 2017. Social Media and Fake News in the 2016 Election. Journal of Economic Perspectives 31, 2 (May 2017), 211–36. DOI: 10.1257/jep.31.2.211

Pepa Atanasova, Isabelle Augenstein, and Christina Lioma. 2020. Diagnostic dataset construction with minimal supervision for interpretable evaluation of fact verification systems. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 1802–1812.

Bilal Ghanem, Simone Paolo Ponzetto, Paolo Rosso, and Francisco Rangel. 2021. Fakeflow: Fake news detection by modeling the flow of affective information. arXiv preprint arXiv:2101.09810 (2021).

Andreas Hanselowski, Hao Zhang, Zile Li, Daniil Sorokin, Benjamin Schiller, Claudia Schulz, and Iryna Gurevych. 2018. A retrospective analysis of the fake news challenge. In Proceedings of the 27th International Conference on Computational Linguistics. 1859–1874.

Beizhe Hu, Qiang Sheng, Juan Cao, Yuhui Shi, Yang Li, Danding Wang, and Peng Qi. 2024. Bad actor, good advisor: Exploring the role of large language models in fake news detection. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 38. 22105–22113.

Ziqing Hu, Yuan Liu, Di Jin, Yucheng Li, Zitao Liu, and Xinyu Lei. 2023. Bad Actor, Good Advisor: Exploring the Role of Large Language Models in Fake News Detection. arXiv preprint arXiv:2309.12247 (2023).

Caio Libanio Melo Jeronimo, Leandro Balby Marinho, Claudio EC Campelo, Adriano Veloso, and Allan Sales da Costa Melo. 2019. Fake news classification based on subjective language. In Proceedings of the 21st International Conference on Information Integration and Web-based Applications & Services. 15–24.

Xinyi Li, Yongfeng Zhang, and Edward C Malthouse. 2024. Large Language Model Agent for Fake News Detection. arXiv preprint arXiv:2405.01593 (2024).

Ye Liu, Jiajun Zhu, Kai Zhang, Haoyu Tang, Yanghai Zhang, Xukai Liu, Qi Liu, and Enhong Chen. 2024. Detect, Investigate, Judge and Determine: A Novel LLM-based Framework for Few-shot Fake News Detection. arXiv preprint arXiv:2407.08952 (2024).

Mohammad Vatani Nezafat and Saeed Samet. 2024. Fake News Detection with Retrieval Augmented Generative Artificial Intelligence. In 2024 2nd International Conference on Foundation and Large Language Models (FLLM). 160–167. DOI: 10.1109/FLLM63129.2024.10852474

University of Chicago. 2024. Set-based (Jaccard) similarity. [link]. Acessado em: 24 de setembro de 2024.

Eleftheria Papageorgiou, Christos Chronis, Iraklis Varlamis, and Yassine Himeur. 2024. A Survey on the Use of Large Language Models (LLMs) in Fake News. Future Internet 16, 8 (2024). [link]

Veronica Perez-Rosas, Bennett Kleinberg, Akira Lefevre, and Rada Mihalcea. 2021. Disinformation and the language of deception: Linguistic features of fake news stories. Digital Threats: Research and Practice 2, 1 (2021), 1–25.

Hannah Rashkin, Eunsol Choi, Jin Yea Jang, Svitlana Volkova, and Yejin Choi. 2017. Truth of varying shades: Analyzing language in fake news and political fact-checking. In Proceedings of the 2017 conference on empirical methods in natural language processing. 2921–2927.

Ali Raza, Yash Paul, Abhik De, and Ashutosh Modi. 2024. Fake News Detection: Comparative Evaluation of BERT-like Models and Large Language Models with Generative AI-Annotated Data. arXiv preprint arXiv:2401.14276 (2024).

Shaina Raza, Drai Paulen-Patterson, and Chen Ding. 2025. Fake news detection: comparative evaluation of BERT-like models and large language models with generative AI-annotated data. Knowledge and Information Systems (2025), 1–26.

Kai Shu, Amy Sliva, Suhang Wang, Jiliang Tang, and Huan Liu. 2017. Fake news detection on social media: A data mining perspective. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. ACM.

Jinjin Su, Xinyi Zhou, and Kai Shu. 2023. Adapting Fake News Detection to the Era of Large Language Models. arXiv preprint arXiv:2311.04917 (2023).

Jinjin Su, Xinyi Zhou, and Kai Shu. 2023. Fake News Detectors are Biased against Texts Generated by Large Language Models. arXiv preprint arXiv:2309.08674 (2023).

Shivani Tufchi, Ashima Yadav, and Tanveer Ahmed. 2023. A comprehensive survey of multimodal fake news detection techniques: advances, challenges, and opportunities. International Journal of Multimedia Information Retrieval 12, 2 (2023), 28.

Lucas Lima Vieira, Caio Libanio Melo Jeronimo, Claudio E. C. Campelo, and Leandro Balby Marinho. 2020. Analysis of the Subjectivity Level in Fake News Fragments. In Proceedings of the Brazilian Symposium on Multimedia and the Web (São Luís, Brazil) (WebMedia ’20). Association for Computing Machinery, New York, NY, USA, 233–240. DOI: 10.1145/3428658.3430978

Theresa Wilson, Paul Hoffmann, Swapna Somasundaran, Jason Kessler, Janyce Wiebe, Yejin Choi, Claire Cardie, Ellen Riloff, and Siddharth Patwardhan. 2005. OpinionFinder: A system for subjectivity analysis. In Proceedings of HLT/EMNLP 2005 Interactive Demonstrations. 34–35.

Jiaying Wu, Jiafeng Guo, and Bryan Hooi. 2024. Fake News in Sheep’s Clothing: Robust Fake News Detection Against LLM-Empowered Style Attacks. In Proceedings of the 30th ACM SIGKDD conference on knowledge discovery and data mining. 3367–3378.

Ruoyu Xu and Gaoxiang Li. 2024. A comparative study of offline models and online llms in fake news detection. arXiv preprint arXiv:2409.03067 (2024).

Xinyi Zhou and Reza Zafarani. 2020. A survey of fake news: Fundamental theories, detection methods, and opportunities. ACM Computing Surveys (CSUR) 53, 5 (2020), 1–40.