Tremendous research has been done in the area of computer security using biometrics. But not much has been done in the field of inverse biometrics, which consists of synthesizing artificial biometric samples that can be used for testing exiting biometric systems or protecting them against forgeries. Due to the complexity of the data collection process and privacy and legal issues that are involved, finding volunteers for data collection is a challenging task. In this thesis, we introduce for the first time an inverse biometrics model for keystroke dynamics that can be used to generate as much data as desired. We show that these synthetic data behave as close as possible like real human data, making our inverse biometrics model a model of choice for testing the existing and upcoming biometric systems. Keystrokes dynamics biometric is a behavioural biometric technology, which allows user recognition based on the actions received from the keyboard while interacting with a graphical user interface. The proposed inverse biometric model first learns from the real human data and based on this experience, it generates synthetic users. Each synthetic user generated by model has a unique behaviour, but follows the properties of real human users. A twofold cross-validation testing technique is employed to validate the synthetic data using a suitable model. Comparable performance results are obtained when applying the model to real human data.