A team of cybersecurity researchers has warned about a new surveillance technique that can identify and track people using ordinary WiFi signals—even if they are not carrying a smartphone or any connected device. The research suggests that standard WiFi routers could potentially become invisible tracking systems capable of recognising individuals with remarkable accuracy, raising fresh concerns about privacy in an increasingly connected world.
Developed by researchers at the Karlsruhe Institute of Technology (KIT) in Germany, the technique uses artificial intelligence to analyse WiFi radio signals and create unique “radio fingerprints” of people moving through an area.
How the Technology Works
The attack, known as BFId, exploits Beamforming Feedback Information (BFI), a feature introduced with WiFi 5 to improve wireless performance. Devices connected to a WiFi network continuously send these feedback signals to routers, and the researchers found that this information is transmitted without encryption.
By collecting these radio signals, machine learning algorithms can analyse how WiFi waves bounce off a person’s body, effectively creating a unique radio-based image that can be used for identification. Unlike traditional surveillance systems, this method does not require cameras, facial recognition, or even a smartphone.
Nearly 100% Identification Accuracy
The researchers tested the technology on 197 participants and reported an identification accuracy of approximately 99.5%. According to the study, the system successfully recognised individuals regardless of their walking style, viewing angle, or whether they were carrying any electronic device.
Once the AI model is trained, identifying a person reportedly takes only a few seconds, making the technology practical for continuous monitoring in areas covered by WiFi networks.
Turning Off Your Phone May Not Help
One of the most concerning findings is that disabling your smartphone or leaving it at home does not necessarily prevent tracking.
As long as nearby devices remain connected to a WiFi network, the router continues exchanging radio signals. These signals interact with people moving through the area, allowing the AI system to detect and identify them without requiring any device to be carried by the individual being tracked.
Why Privacy Experts Are Concerned
Unlike CCTV cameras, which are visible and easily identifiable, WiFi networks already exist in homes, offices, airports, shopping malls, cafés, and public spaces. Researchers warn that widespread adoption of this technology could create an almost invisible surveillance infrastructure that operates without attracting attention.
They caution that such systems could potentially be misused by governments, businesses, or malicious actors to monitor people’s movements and identities without their knowledge or consent.
Researchers Call for Stronger Safeguards
The KIT research team is urging industry bodies to introduce stronger privacy protections into future WiFi standards, particularly the upcoming IEEE 802.11bf specification.
Experts recommend encrypting Beamforming Feedback Information and implementing additional security measures to prevent unauthorised collection of radio signal data. While the technology is still primarily a research demonstration, the findings highlight the growing privacy challenges posed by advances in artificial intelligence and wireless networking.
Summary
Researchers from Germany’s Karlsruhe Institute of Technology have demonstrated a WiFi-based tracking method that can identify people without requiring them to carry a smartphone or connected device. Using unencrypted Beamforming Feedback Information and AI, the system achieved nearly 99.5% identification accuracy in testing. The researchers warn that ordinary WiFi networks could become invisible surveillance tools unless stronger privacy protections are incorporated into future wireless standards.
