Fake news refers to false information or misinformation, and its rapid spread can lead to significant consequences. Often, distinguishing between fake news and real news becomes challenging, resulting in counterfeit news gaining more traction and going viral faster than genuine news. Recognizing the urgency of this issue, this study has been undertaken to detect fake news effectively. The research analyzes a news dataset, leveraging an advanced technique known as Particle Swarm Optimization for feature selection. PSO, an evolutionary algorithm, offers a novel approach to feature selection, moving beyond the limitations of traditional methods or their hybrid variations, which have been commonly used thus far. Once the features were selected, they were classified using six machine learning algorithms. The performance of these models was rigorously evaluated using standard evaluation metrics. The Support Vector Machine (SVM) demonstrated the highest performance among all the algorithms, outperforming the other classifiers in detecting fake news. This study introduces innovation in feature selection and highlights the potential of evolutionary algorithms in enhancing the reliability of fake news detection systems.