To enhance the interference detection and emergency response capabilities of civil aviation radio monitoring systems, an adaptive blind GNSS interference detection method was proposed based on random matrix theory. The maximum eigenvalue detection(MED) algorithm in the field of spectrum sensing was transplanted to the civil aviation GNSS interference detection for the first time, and the threshold value of the traditional MED algorithm was improved. The adaptive maximum eigenvalue detection (A-MED) algorithm was proposed, which significantly improved the detection performance under the condition of low signal-to-noise ratio without prior information. The simulation experiments show that the A-MED algorithm has the best performance in the detection experiments of single-tone, multi-tone, sweep frequency, pulse and some noisy interference signals. The detection probability has reached 0.8 at a low signal-to-noise ratio (-30 dB) and is rapidly approaching 1. For pulse interference (PI), the energy detection method (ED) has limitations when the pulse amplitude is small, while both MED and A-MED can effectively detect it.