This paper proposes a new hybrid ML model that relies on both K-Means clustering and the Variational Bayesian Gaussian Mixture model to efficiently detect unknown network attacks. The proposed model first classifies the input data into various clusters using K-Means. Then, it identifies anomalies in those clusters using the Variational Bayesian Gaussian Mixture model, to be then classified as unknown. The proposed model shows promising results when identifying whether a data point is an attack or not with an F1 score of up to 91\%, such that the Variational Bayesian Gaussian Mixture model detected up to 86\% of unknown attacks. The conducted experiments shows acceptable performances, where the predictive pipeline took around 2.42 seconds to be processed.