Smart building incorporates the ubiquitous sensing ability of Internet of things (IoT) technology with the industrial infrastructure for automating decision‐making. Inspired by this, a novel technique of stochastic game net (SGN) has been proposed to minimize the tangible and intangible infrastructural losses due to disasters. Specifically, IoT technology is used for disaster management and control in a smart building based on SGN. In the proposed technique, every IoT sensor act as an individual player with prefixed action sets and strategies. Complete SGN of the IoT network in the smart building is formalized by collaborating SGN of an individual sensor. The model deployed in IoT building with different sensors detects disasters beforehand and generates early warning alerts to management units based on game‐theoretic decision‐making. Experimental evaluation on four challenging datasets, namely, disaster/accident sources, gas sensor drift, red cross smoke alarm, and oil spills show the effectiveness of a proposed framework in the smart building. The results are compared with several state‐of‐the‐art decision‐making techniques to measure the overall effectiveness of the proposed system. Based on simulations performed on NetLogo 5.3.1, enhanced results of performance parameters including temporal delay, statistical efficiency, reliability, and stability were registered. Moreover, mathematical analysis has been carried out to determine the overall performance of the presented technique.