The study proposes an innovative approach to enhance energy efficiency in Wireless Sensor Networks (WSNs) for smart city applications. The primary focus is on leveraging distributed artificial intelligence (AI) and multipath routing techniques to address challenges such as unequal clustering, poor cluster head selection, and excessive power consumption within WSNs. The approach uses agent-based clustering, where autonomous AI agents dynamically form clusters of sensor nodes based on real-time data characteristics. These clusters are then used for multipath routing, optimizing energy consumption, reliability, and congestion reduction. The distributed nature of AI agents allows for adaptive cluster formations. This algorithm aims to address issues related to uneven clustering, inefficient cluster head selection, and excessive power consumption. Additionally, the integration of agent-based clustering is proposed, involving the deployment of autonomous AI agents that dynamically cluster sensor nodes based on real-time data properties. These AI agents facilitate self-organization and adaptability, ensuring that clusters accurately reflect the evolving data landscape in urban environments. The approach also employs sophisticated energy management strategies at the sensor node level, such as duty cycling, adaptive transmission power control, and sleep-wake scheduling. Simulations in a smart city environment show significant improvements in energy efficiency, prolonging the network’s operational lifespan and improving service quality by mitigating data loss and latency issues. This approach contributes to the sustainable development and performance optimization of smart city infrastructure.