The intrusion detection system plays an important role in securing our system, by preventing our system from intruders. However, traditional intrusion detection, such as user authentication, encryption, and firewall have failed to completely protect networks and systems from the increasing and sophisticated attacks and malwares. The presented new method classifies network behaviour as normal or abnormal while reducing misclassification. Ant Colony Optimization (ACO) algorithms can be applied to the data mining field to extract a set of rules for detection and classification. Support Vector Machine (SVM) is a technique for detecting intrusions in the system, which can provide real-time detection capability and it can deal with large dimensionality of data. SVM can learn a larger set of patterns and be able to scale better because the classification complexity does not depend on the dimensionality of the feature space. In this paper, Active learning Support Vector Machine and Ant Colony clustering are combined to detect the network intrusion. Combining SVM and Ant Colony (CSVAC) uses both the algorithm while avoiding their weaknesses. This algorithm is implemented and evaluated using standard benchmark KDDCUP99 data set.