<p>Social networks are globally networked platforms that enable information sharing and worldwide communication. However, with growing usage of social network, there is a considerable increase in the number of automated accounts called malicious social bots. Social bot detection is challenging due to ambiguous features of accounts, class imbalance, and behavioral similarity between genuine and bot accounts. For instance, these bots usually mimic human communication, activity pattern and are primarily created with maligned intent for spreading false information, and causing chaos in the society. In order to protect integrity and reliability of online interactions, the field of social media bot detection has become a crucial area of study. Our research introduces an innovative approach, namely Twitter/X Bot-detection using explainable artificial intelligence, namely TwiBotX, to identify social bots using machine learning with 14 profile features. Our study utilizes the benchmark TwiBot-22 database for model building and validation. The investigation employs hyperparameter optimization through Grid Search technique which improves the accuracy of model. The experimentation findings demonstrate that our model TwiBotX distinguishes social media bots from human accounts with an accuracy of 96.47% and average model precision of 0.99. Our model shows better performance results when compared to different existing bot detection benchmarks. We have also incorporated explainability using shapley additive explanations model to identify essential features which drive bot detection results through explainability analysis. The validation of our results involved comparing TwiBotX with multiple existing API standards. The performance of TwiBotX exceeds that of multiple existing social bot detection systems. The system protects TwiBotX from automated cyberattacks which safeguards the platform integrity.</p>