<p>Social Networks (SNs) are essential for communication and information exchange in modern life, with platforms like Facebook, Twitter, and LinkedIn significantly influencing personal interactions and business operations. A Wireless Sensor Network (WSN) can be modeled as a SN, with sensor nodes analogous to individuals or organizations and communication links resembling SNs. Data transmission in WSNs relies on communication pathways between sensor nodes, similar to SNs where information flow depends on user friendships. However, the reliability of such networks is highly dependent on the presence of certain key nodes. Failure of these nodes can lead to network disconnection; such nodes are referred to as critical nodes (CNs) or cut vertices. Identifying these nodes is essential for ensuring network reliability, connectivity, stability, functionality, robustness, and resilience. This paper introduces a novel bio-inspired metaheuristic algorithm called Revitalized Boltzmann Distribution-based Grey Wolf Optimization (RBD-GWO) to optimize CN detection (CND). RBD-GWO initializes a population of Grey Wolves and evaluates their fitness using a Boltzmann-based pairwise connectivity measure. Based on this evaluation, the wolves are ranked and iteratively updated through the optimization process, with the wolf exhibiting the highest fitness selected as the CN. The performance of RBD-GWO is validated using datasets from the Stanford Large Network Dataset Collection. The results demonstrate that RBD-GWO achieves higher detection accuracy, lower error rates, and reduced complexity compared to conventional techniques. Specifically, the method attains an 8% improvement in accuracy, along with reductions of 38%, 13%, and 11% in error rate, detection time, and space complexity, respectively. These results underscore the effectiveness and reliability of the RBD-GWO for CND in various network applications.</p>