The exponential increase in internet users has led people to start using technology-based development for carrying out activities against the law. According to the studies made from previous knowledge and experiences it can be found out that, cyber terrorism puts out stress and anxiety, increases feelings of intrusion, threat and hardens the political frame of mind. Therefore, it is important to identify the fact that cyber terrorism has had a very awful and shocking effect on our society. Thus, With the use of a Machine learning model, it helps us to predict the possible future terrorist activities by detecting the instigating text messages in social network platforms. The proposed system aims to develop a method for of prediction of future terrorist activities using a machine learning algorithm that will help the system in detecting potential terrorist attacks and in providing safeguards and strengthening defense for required areas consequently lessens the loss of life and property. Along with that it also provides Fast detection to illegal activity, improving information accuracy. It has the ability to distinguish between instigating (terrorism-related) messages and non-instigation (non-terrorism-based) messages. This method also helps in identifying the user responsible for sending such messages and also recognize the specific messages and also shows the best machine learning algorithm that best fits and that can be used for prediction based on the accuracy analysis of the algorithms.