Hidden danger investigation is a crucial method for ensuring the safety of air traffic control, and the graded assessment of hidden dangers serves as an important basis for their management. Focusing on the hidden dangers associated with air traffic control, a grading model was established based on cloud theory grounded in expert evaluations. In response to the relatively coarse categorization provided by existing regulations regarding hidden dangers, a more precise evaluation framework was developed to enhance the accuracy of assessments. To improve the reliability of these evaluations, a group decision-making approach involving experts was employed for assessing hidden danger levels. Furthermore, to enhance the objectivity of the evaluation process, a cloud model was introduced for quantitative assessment alongside qualitative analyses. Finally, an illustrative example was presented to demonstrate the evaluation procedure. The research findings indicate that the cloud model-based approach exhibits strong feasibility and applicability.