A feature selection algorithm examines the data to weed out distracting, extraneous information while improving classification accuracy. The finest feature subgroup for classification purposes is chosen in this work using the grey wolf optimization (GWO) technique. The newest bio-inspired optimization methods, such as the grey wolf optimizer (GWO), imitate grey wolves' natural leadership structure and hunting strategy. The intelligent grey wolf optimization seeks out the best sections of the composite search space through the interaction of individuals in the community. As a result, selecting features is considered to be used in pre-processing before smearing a classifier to a data set. The feature selection strategy effectively maximizes classification accuracy and reduces computing costs. The developed system is tested using KDD CUP 1999 data sets. The experimental outcome displays a comparison of the IDS data set's accuracy, sensitivity, and specificity when employing various categorization approaches. In order to make computer system infiltration accessible, the key goal is to choose the smallest number of characteristics that will be ideal and give high classification accuracy of big datasets.