Огляд штучних нейронних мереж в контексті створення Deepfake

Abstract

The article examines modern approaches to the creation and use of Deepfake technology in the context of information attacks. The main methods of forming audiovisual content using artificial intelligence tools are analyzed. A classification of typical image and video manipulation techniques is presented, and common approaches to their implementation are characterized. The paper focuses on the growing risks associated with the use of Deepfake to spread disinformation, undermine trust in digital content, and interfere with personal identity. The author emphasizes the relevance of further development of ef ective means of detecting fakes and improving methods of protection against such threats.