The development of modern electric power systems is associated with digitalization and an increase in the data flow from electric power facilities to the control centers. On the other hand, an increase in the renewable energy sources share leads to an increase in electrical modes uncertainty and a decrease in total inertia, which imposes new requirements on the speed of the procedure for assessing dynamic stability and emergency control. The application of traditional deterministic algorithms to the analysis of the dynamic stability of power systems in the presence of increased requirements for performance may turn out to be inefficient. To overcome the shortcomings of traditional methods for assessing the dynamic stability of power systems, artificial intelligence methods can be used. This class of methods has a significant speed of trained models and the ability to search for patterns in the data, which makes it effective in modern power systems. The paper presents the results of developing a method for assessing the dynamic stability of a power system based on artificial intelligence methods, taking into account the topological connectivity of the electrical network. The technique is based on the application of the gradient boosting algorithm for decision trees. Numerical simulation was performed on the IEEE39 model implemented in Matlab/Simulink; the Scikit-learn library of the Python3 programming language was used to implement machine learning algorithms. To train the machine learning algorithm the load angles of synchronous generators, voltage levels at the nodes connecting synchronous generators to the electrical network, the topology of the electrical network, the duration and resistance of the short circuit were used. As a result of applying the trained algorithm, taking into account the topology of the electrical network, an accuracy of 91.5% was obtained on the test sample. The accuracy on the test sample without taking into account the topological connectivity of the elements of the power system was 81.6%.