Beyond speculation: Measuring the Growing Presence of LLM-generated texts in Multilingual Disinformation

Large Language Models (LLMs) have demonstrated unprecedented capabilities in generating human-like text, raising significant concerns about their potential misuse in creating disinformation campaigns across multiple languages. This study moves beyond speculation to provide empirical measurements of the growing presence of LLM-generated texts in multilingual disinformation ecosystems. We analyze the proliferation of AI-generated content across diverse linguistic contexts, examining how different languages and cultural contexts influence the detection and spread of LLM-generated disinformation. Our findings reveal concerning trends in the sophistication and scale of AI-generated multilingual disinformation, highlighting the urgent need for robust detection mechanisms and cross-linguistic approaches to combat this emerging threat.

In IEEE (Accepted)