The proliferation of fake news on social media poses a significant challenge, necessitating robust detection systems combined with explainability to ensure trust and transparency. This paper explores how sophisticated machine learning techniques can be integrated with Explainable Artificial Intelligence (XAI) to address the dual challenges of detecting fake news and explaining the findings. We propose an innovative framework for explainable fake news detection, leveraging the capabilities of BART for both classification and automated rationale generation, complemented by SHAP for granular feature importance analysis. Recent studies have successfully applied BERT and SHAP for explainable fake news detection in domains such as COVID-19 misinformation [11], highlighting the importance of understandable insights, which form a core mechanism of our proposed system. Our experiments on real datasets, such as LIAR and ISOT, show that our suggested framework detects fake news considerably more accurately while providing useful explanations for the identified fake news. The results indicate that the integration of BART and SHAP not only enhances fake news detection but also offers comprehensive insight into the decision-making process, thus making it a viable solution in the fight against misinformation in the current digital age.