Researchers at the Karlsruher Institut für Technologie (KIT) in Germany have demonstrated a new surveillance method that uses ordinary Wi-Fi signals and artificial intelligence to identify individuals with near-perfect accuracy. This passive tracking method works even if the person being monitored is not carrying any active wireless device, raising fresh concerns about the privacy of everyday networks.
Instead of relying on specialized sensors or expensive hardware, the tracking system exploits normal communication between Wi-Fi routers and connected client devices. It utilizes unencrypted Beamforming Feedback Information (BFI), which devices regularly transmit to routers to optimize signal direction. Because BFI is transmitted without encryption under current standards, anyone within physical range of the radio waves can capture these signals. After training the machine learning model, identifying a specific person takes only a few seconds.
According to the research paper, titled "BFId: Identity Inference Attacks Utilizing Beamforming Feedback Information," the team tested the system with 197 participants. The machine learning model analyzed the unique ways radio waves bounced off the subjects' bodies as they moved. Even when participants changed their walking angles, the system identified them with nearly 100 percent accuracy.
Professor Thorsten Strufe, a cybersecurity expert from KIT's KASTEL Institute of Information Security and Dependability, explained that the technology functions similarly to a traditional camera but uses radio waves instead of light waves. Because nearby wireless devices continuously generate signal activity, turning off a personal smartphone is not enough to prevent detection.
Julian Todt, a researcher at KASTEL, warned that the widespread presence of Wi-Fi networks in offices, cafes, and public areas could allow businesses or public authorities to track citizens invisibly. Fellow researcher Felix Morsbach noted that while intelligence agencies and criminals already exploit hacked security cameras and smart doorbells, Wi-Fi networks present a unique threat because they are ubiquitous and completely hidden from view.
The KIT researchers are calling for stronger privacy protections to be integrated into the upcoming IEEE 802.11bf standard, which is currently being developed to regulate Wi-Fi sensing technologies. The research team originally presented its findings at the ACM Conference on Computer and Communications Security (CCS) in November 2025.
