Vishing Detection and Prevention in Modern Communication Systems: A Systematic Literature Review

Vishing, or voice phishing, has evolved from a basic scam to a sophisticated threat that exploits VoIP technology, mobile banking, and artificial intelligence for creating convincing synthetic voice clips.This study reviews 25 works on vishing detection and prevention solutions.Text-based solutions, based on natural language processing techniques, have almost perfect performance in offline call transcript detection and processing similar SMS messages.Audio and multimodal techniques widen the detection scope to noisy recordings, synthetic speech, and audio deepfakes, employing convolutional-BiLSTM architectures and selfsupervised deep learning models.Human-oriented vishing detection techniques show how social, demographic, and personalized aspects can increase users' resistance against vishing attempts.Recent research into voice modulation and audio deepfakes demonstrates how vishing attacks continue evolving to produce realistic voices, requiring more efficient anti-spoofing tools.Knowledge gaps within this literature include a lack of standardized vishing datasets, evaluation of solutions in deployment, comprehensive voice threat detection, and integration of behavioral and technological measures.