The advancements in the distributed energy system and digital technology in the smart grid system increase efficiency, stability, and reliability. However, it increases the vulnerabilities in the grid network. The falsely injected data in the grid network leads to failures in energy production, and consumption. Hence, an Ant Lion-based Modular Neural network (ALbMNN) model was proposed to detect the normal and malicious data. The presented model integrates the ant lion fitness and MNN attribute to detect the falsely injected data in the grid system. The dataset was initialized and pre-processed using the divide-and-conquer principle of MNN. The optimal ant lion fitness solution helps in selecting features optimally from the dataset. Finally, the presented model was assessed with a large-scale smart grid dataset, and the results are estimated. Moreover, a comparative analysis was performed to verify the performance of the developed scheme. Based on performance and comparative analysis, the suggeated model performed better than other existing methods.