Abstract Today, ‘security challenge’ is considered a commonly‐used catchword when it comes to emerging technologies such as the internet of things (IoT). Turning a blind eye to this challenge can sometimes lead to irreparable human and financial damage in everyday life. Threat modelling, despite being time consuming, complex, and error‐prone, is still recognized as a preventative solution in the design phase. The method proposed in this paper, called automatic generation of threat paths, is suggested as an automated solution to the problem of identifying and determining potential threats. The proposed method is the improved and robust development of automated modelling and the introduction of the integration of new features (such as conditional probability and security) with the techniques of previous generations (such as Petri Nets (PNs)). The proposed method is evaluated using different security scenarios, and the results show that the proposed method outperforms other manual methods in this domain in terms of both time and cost (99.9% and 98.05%, respectively).