The paper examines the key features of deepfakes and approaches to their recognition using computer vision and machine learning methods. In the course of the study, the signs of deepfake detection were identified and analyzed. Based on priorities, the features that ensure high recognition accuracy were identified and a conclusion was drawn about the significance of each feature.

Keywords: generative artificial intelligence, disinformation, deepfake, deepfake detector, cybersecurity, fraud, recognition features, analysis, recognition vectors, machine learning, model