This study aims to sense trust and distrust in a real-time inspired scenario through the classification of brain signals. Here, a word elicitation study is used to invoke the mental state associated with trust and distrust associated with the machine. Participants think of any event or experience that comes into their mind when they observe the word. They think or recall that event/experience without deliberately filtering out any kind of cognitive or affective mental state, which we consider as a replica of a real-life scenario where all kinds of mental states or emotions possibly co-exist along with trust or distrust. While thinking or recalling such events, Electroencephalography data is recorded from the participants' cortex and analyzed through Machine Learning approaches with several classification algorithms. The study developed an approach to sense whether the human is going through trust or distrust and compared different methods to discuss their efficacies in different scenarios. Here, the individualistic and generalistic approach is delved into, and it found that individualistic approaches provide better accuracy in sensing trust or distrust state of the human brain. Also, this study explored ways to increase the efficiency of the method by reducing the number of channels and compared the performance of the models by observing the loss of accuracy caused by the reduced number of channels. This study found that the K-Nearest Neighbor and/or Random Forest classifier algorithm provides the best result using raw data with the individualistic approach in most scenarios, achieving up to 100% average accuracy.