REVIEW OF EXISTING METHODS FOR ASSESSING DISINFORMATION RISKS IN THE CONTEXT OF HYBRID WARFARE

The article is devoted to analyzing modern approaches to assessing disinformation risks in the context of hybrid warfare. The primary technologies for automated fake news detection are reviewed, key gaps in the approaches are identified, and prospects for further research are proposed. The review confirms that existing models are still insufficiently adapted to the rapid changes in disinformation tactics, including “adversarial AI” techniques and dynamic shifts in fake narratives. Key issues hampering the development of effective systems for combating fake news include the lack of localized datasets in multiple languages, insufficiently defined legislative norms, and a lack of interdisciplinary approaches that integrate psychological and social aspects of perception of manipulative messages. At the same time, research has shown that combining technological methods (machine learning, social media analysis, multicriteria risk assessment) with expert and user input can significantly improve the accuracy and speed of identifying fake news while prioritizing response measures. The conclusions outline prospects for further developments, including multimodal detection systems capable of a