The airborne warning system applied on modern civil aviation aircraft can reduce flight safety issues caused by human factors to a certain extent. However, the current warning systems installed on aircraft are not sufficient to merely warn the pilots, because the plane may be hijacked or maliciously operated, and the pilots themselves may choose to ignore the warnings due to reasons such as aversion to the warning system or fatigue. Therefore, adding human-machine authority management functions oriented towards human-machine mixed decision-making to the civil aircraft's airborne system has become a hot topic in civil aircraft flight management research. Firstly, based on the flight characteristics of civil aircraft, this paper designs the overall framework of a dynamic management system for human-machine authority in civil aircraft. Secondly, the method of predicting the deviation of the flight route is adopted to determine the flight status of the aircraft, and a method based on power spectrum estimation is proposed to judge the fatigue level of the pilot. Additionally, a credibility model for the pilot is proposed to assess their instant driving ability. Based on the above research, a Mamdani-type fuzzy logic reasoning system is designed to achieve dynamic adjustment of human-machine authority. Finally, a verification program for the dynamic management system of human-machine authority in the cockpit of civil aircraft is developed, and simulation verification is carried out under typical flight scenarios.