Disclosed herein is an artificial intelligence device including a communication interface configured to receive a reference signal in a time domain for echo cancellation of an external device and a processor configured to determine whether a spectral density at at least one test target frequency is less than or equal to a preset spectral density reference value at at least one test target time point and determine that a failure occurs in an audio component when the spectral density is less than or equal to the preset spectral density reference value.