This study presents a comparative analysis of fault detection and classification systems developed for current sensors in a drive system with a permanent magnet synchronous motor (PMSM). The research focuses on a fault detection algorithm based on current signal markers and on classification systems employing both shallow and deep neural network architectures. The study aims to evaluate the effectiveness, accuracy, and robustness of these methods in identifying sensor faults that may influence the performance and stability of the drive control system. Experimental verification was carried out using the same Moog 0.894 kW motor, tested under various load and speed conditions to ensure reliable comparison and validation of the obtained results.