This research analyzes the capabilities of adaptive filtering for suppressing Electromagnetic Interference in digital signals characteristic of automotive onboard networks. The behavior of three distinct adaptive algorithms, namely Least Mean Squares, Normalized Least Mean Squares, and Intelligent Normalized Least Mean Squares, integrated with an intelligent interference detection system, is investigated. A realistic test environment is created, in which the useful signal is subjected to two critical types of interference: impulsive and harmonic, as well as background Gaussian noise. A modified solution is proposed, employing an interference detector based on the analysis of the Mean Squared Error and signal kurtosis, which dynamically controls the adaptation step of the Intelligent Normalized Least Mean Squares algorithm. It is established that Intelligent Normalized Least Mean Squares exhibits maximum effectiveness in the presence of harmonic interference. It is revealed that even standard Normalized Least Mean Squares and Least Mean Squares filters demonstrate a significant reduction in the Mean Squared Error compared to the initial noise level; however, Normalized Least Mean Squares proves to be more stable than Least Mean Squares due to normalization. The advantage of Intelligent Normalized Least Mean Squares over fixed algorithms is substantiated when operating with impulsive interference, which requires rapid adaptation. Numerical results and time graphs are presented, confirming the obtained results.