Researchers in Germany have demonstrated that standard Wi-Fi signals can be leveraged to identify people without the use of cameras. The method is based on recording how radio waves change within a space, achieving an identification accuracy of 99.5% in an experiment involving 197 participants under controlled conditions.
How Wi-Fi Becomes a 'Sensor'
This technique, developed by the KASTEL Institute of Information Security and Dependability at the Karlsruhe Institute of Technology, functions similarly to echolocation. As signals travel through a room, the changes caused by a person's presence reveal their outline, position, and movement. By processing these wireless signals through AI systems, a representation is created that is sufficient for distinguishing between different individuals.
At the heart of the method lies Beamforming Feedback Information (BFI). This refers to data exchanged between devices and a router to optimize signal direction. Researchers found that this information can be analyzed to extract details about people moving through the signal's path, using standard Wi-Fi hardware rather than specialized, expensive equipment.
An Invisible Threat to Privacy
One of the most concerning findings of the study is that turning off one's personal smartphone is not necessarily enough to prevent tracking. The technique can function as long as other active devices in the same space are communicating via Wi-Fi. BFI data is transmitted without significant protection against interception and can be collected by anyone with the necessary technical skills within the network's range.
Unlike cameras, which are physically detectable, radio waves are invisible and already omnipresent in homes, offices, and public spaces. Researchers warn that without built-in safeguards, widespread wireless networks could be transformed into mass surveillance infrastructure, posing risks to vulnerable groups such as journalists and political dissidents. Consequently, they are calling on manufacturers and standardization bodies to integrate privacy-by-design into future Wi-Fi protocols.
