Publication Details
Disinformation Capabilities of Large Language Models
Pikuliak Matúš
Srba Ivan
Móro Róbert
Macko Dominik
Bieliková Mária, prof. Ing., PhD. (DCGM FIT BUT)
large language models, disinformation generation, human evaluation, fake news detection
Automated disinformation generation is often listed as one of the risks of large language models (LLMs). The theoretical ability to flood the information space with disinformation content might have dramatic consequences for democratic societies around the world. This paper presents a comprehensive study of the disinformation capabilities of the current generation of LLMs to generate false news articles in English language. In our study, we evaluated the capabilities of 10 LLMs using 20 disinformation narratives. We evaluated several aspects of the LLMs: how well they are at generating news articles, how strongly they tend to agree or disagree with the disinformation narratives, how often they generate safety warnings, etc. We also evaluated the abilities of detection models to detect these articles as LLM-generated. We conclude that LLMs are able to generate convincing news articles that agree with dangerous disinformation narratives.
@INPROCEEDINGS{FITPUB13150,
author = "Ivan Vykopal and Mat\'{u}\v{s} Pikuliak and Ivan Srba and R\'{o}bert M\'{o}ro and Dominik Macko and M\'{a}ria Bielikov\'{a}",
title = "Disinformation Capabilities of Large Language Models",
pages = "14830--14847",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
year = 2024,
location = "Bangkok, TH",
publisher = "Association for Computational Linguistics",
ISBN = "979-8-8917-6094-3",
doi = "10.18653/v1/2024.acl-long.793",
language = "english",
url = "https://www.fit.vut.cz/research/publication/13150"
}