Scientists at Karlsruhe Institute of Technology have developed an experimental system that can identify individual people using nothing more than ordinary WiFi signals. Unlike facial recognition or smartphone tracking, the technology doesn’t rely on cameras, wearable devices, or an app installed on the person being identified. Instead, it analyzes how a person’s body interacts with radio waves already traveling through the environment.

The system, known as BFId (Beamforming Feedback Identification), was created by researchers at KASTEL – Institute of Information Security and Dependability, led by Thorsten Strufe and his colleagues. In experiments involving 197 participants, the researchers reported identification accuracy of up to 99.5% under controlled conditions. According to the team, the system remained highly accurate even when people approached from different directions or changed the way they walked.

Rather than using images, BFId relies on beamforming feedback information (BFI)—technical data that WiFi-enabled devices exchange with routers to optimize wireless connections. As WiFi radio waves bounce off a person’s body, they produce subtle and distinctive reflection patterns. By training an AI model on these patterns, the researchers found they could distinguish one individual from another with remarkable accuracy, effectively creating a form of radio-frequency biometric identification.

One of the study’s most striking findings is that your own phone doesn’t necessarily need to be transmitting. The system only requires nearby WiFi traffic generated by other devices in the environment. That means switching your phone off or enabling airplane mode may not fully prevent this type of identification if other WiFi-enabled devices are communicating with the router.

Professor Strufe has emphasized that the research is intended to highlight a privacy risk, not to encourage surveillance. He warned that, if left unaddressed, future WiFi infrastructure could unintentionally allow routers to become tools for passive monitoring. The researchers are therefore calling for stronger privacy safeguards to be incorporated into the emerging IEEE 802.11bf WiFi sensing standard before the technology becomes widely deployed.