Beyond Speculation: Measuring the Growing Presence of Large Language Model-Generated Texts in Multilingual Disinformation

Download PDF
Open Webpage

Dominik Macko, Aashish Anantha Ramakrishnan, Jason Samuel Lucas, Róbert Móro, Ivan Srba, Adaku Uchendu, Dongwon Lee

Published: 2026, Last Modified: 17 Mar 2026Computer 2026EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Increased sophistication of large language models (LLMs) and the consequent quality of generated multilingual text raises concerns about potential disinformation misuse. Our study provides the first empirical evidence of LLM presence in real-world disinformation data sets.