This study presents a bibliometric study of scientific output on Artificial Intelligence (AI) and disinformation indexed in the Scopus database between 2020 and 2024. The purpose is to identify the most relevant research trends, collaboration networks, prominent authors and institutions, as well as emerging topics in this field. The study is based on a structured search of articles, from which data is obtained using co-authorship techniques, keyword co-occurrence, and citation analysis with the help of VOSviewer software. Among other findings, there has been a consistent increase in publications over the last 2 years, parallel to the rise of commonly used generative AI models and the proliferation of digital disinformation campaigns. Similarly, fragmented co-authorship networks were identified, with low inter-institutional collaboration, except for small, consolidated clusters in countries such as the US, Germany, Spain, and Italy; however, most institutions operate in isolation and cooperation between authors is still limited. Thematic mapping grouped the research into clusters focused on the automated detection of fake news, the risks of deepfakes, and the social and ethical implications of AI-generated disinformation. It is concluded that the field is expanding but still faces challenges regarding scientific integration, global collaboration, and the development of a shared interdisciplinary agenda.