On October 17, 2025, researchers from the KASTEL Institute of Information Security and Reliability at the Karlsruhe Institute of Technology in Germany published a study. The findings indicate that attackers can leverage existing Wi-Fi signals to identify individuals even when they are concealed behind walls. This can be achieved without any physical interaction with the individuals' devices, through the passive observation of alterations in radio wave propagation. Remarkably, tracking remains feasible even when phones are powered off or disconnected from the network.

This technology does not rely on specialized hardware. Instead, it attains highly accurate identity recognition by scrutinizing minute fluctuations in Wi-Fi signals induced by human movement. It then constructs distinctive human motion profiles using machine learning models. The study disclosed that in an experiment involving 197 participants, identity recognition accuracy neared 100% with merely a few seconds of signal analysis. Moreover, this accuracy was impervious to variations in observation angles or walking styles. The relevant paper has been published in the journal ACM Data Science. In light of these findings, experts are urging for enhanced security measures in wireless environments to mitigate the rising threat of surveillance.