Strengthening Digital Security in MIS: A Business Analytics Approach to Deepfake Detection

Digital content security is highly impacted due to the dominance of the deepfake, especially in MIS. This paper explores the combination of business analytics methods to discover and counter deepfake-related cyber threats. What we eventually aimed for was crafting and deploying a model to spot deepfake behavior in network traffic. Our work followed a quantitative research method; we trained and evaluated our model using the CICIDS2017 and UNSW-NB15 datasets. The model performed well in Botnet ARES and Web Attack types and had a precision and recall accuracy of 96%. ‘Packet Length’ and ‘Bwd Packet Length Log’ were statistically significant features for deepfake detection in the feature significance analysis. The confusion matrix also revealed low misclassification, indicating the stability of the model. Based on these findings, using state-of-the-art machine learning models, MIS can be applied to enhancing real-time threat detection to boost organizational cyber security.