This research uncovers that wireless charging of mobile phones generates a leakage magnetic field in the surrounding space, containing information about charging power. Using longwave radio technology with a ferrite rod antenna, signal detection extends to 1.5 meters, enabling inconspicuous measurement. Extracting features from the magnetic field waveforms, machine learning trains a Deep Neural Network model to distinguish user’s activities on the phone with 100% accuracy. This highlights the privacy risks posed by this wireless charging technology, providing valuable insights into potential privacy concerns.