The progress of Web 2.0 has radically transformed how we interact in the present day. Although it has enhanced connectedness and interpersonal communication, it has its disadvantages including sharing of fake information and the effects of the same on mental health. With the current influence of social media in our lives, it is important to thoughtfully weigh the benefits that it has brought to us against the potential harm that it can cause. This research undertaking attempts to examine the sentiment aroused by fake news shared in leading global social media sites like Twitter (X). A hard voting ensemble classifier was used, and it was able to classify tweets into phony and authentic with a remarkable 94.21% accuracy rate. Moreover, sentiment analysis with the same classifier resulted in an accuracy rate of 84.11%. The psychological implications of sharing misinformation with the general population were then examined using these models. The research also considers the significant correlation between fake news and emotions and depicts that misleading information is directly associated with negative mental health conditions. This establishes a strong relationship between the false news and attitudes. The results of the proposed approach were compared to the traditional machine learning algorithms. A larger multilingual dataset could justifiably be considered an improvement of the future study to test false tweets and real-time human emotions. As generative AI and large language models (LLMs) that can create highly realistic and misleading content, such as deepfakes; grow more popular, the issue of emotional reactions to false information has gained even greater significance.