The rapid growth of electrical power demand in India has increased the complexity of transmission networks. Conventional fault detection techniques are often slow in identifying abnormal conditions in modern smart grids. This paper presents an AI-based fault detection model using Ensemble Learning techniques for Indian transmission systems. The proposed system combines Random Forest, Gradient Boosting, and Decision Tree algorithms to improve fault classification accuracy and reduce detection time. Different transmission line fault conditions such as single line-to-ground fault, line-to-line fault, double line-to-ground fault, and three-phase fault are analyzed using simulated grid data. Performance parameters including accuracy, precision, recall, and fault detection speed are evaluated. Results show that the ensemble model provides higher reliability and better fault prediction performance compared to individual machine learning models. The proposed method can support real-time smart grid monitoring and improve transmission system stability in Indian power networks.