KIT Researchers Identify People Through Wi-Fi Routers With 99.5% Accuracy

Researchers at KIT developed BFId, a system that identifies people through standard Wi-Fi routers with 99.5% accuracy using unencrypted beamforming data.

By Marek Novák

Published May 25, 2026 3:06 PM

KIT Researchers Identify People Through Wi-Fi Routers With 99.5% Accuracy

Researchers at the Karlsruhe Institute of Technology (KIT) have developed a system called BFId that identifies individuals using unencrypted beamforming data from standard Wi-Fi routers. The technology achieves a 99.5% identification accuracy rate for people walking through a room. This method requires no specialized hardware and functions even when targets lack Wi-Fi devices.

New system uses unencrypted beamforming data to track individuals in rooms.

BFId outperforms traditional Channel State Information (CSI) methods, which achieved an 82.4% accuracy rate in the same tests. The system processes each data point containing 740 features compared to only 212 for CSI. Researchers attribute this higher spatial resolution and noise filtering from compression as key factors in its superior performance.

BFId system achieves 99.5% identification accuracy using beamforming data with 740 features per point versus 212 for CSI methods.

Professor Thorsten Strufe from KASTEL, KIT's cybersecurity institute, warned that the technology poses significant risks to fundamental rights, particularly privacy. The research team argues that the upcoming IEEE 802.11bf standard for Wi-Fi sensing lacks adequate privacy protections before widespread deployment. They are calling for new safeguards to be added to the standard.

Mitigation strategies such as reducing report frequency have minimal effect on preventing identification. Encrypting beamforming data would likely break backward compatibility with existing devices, creating a technical hurdle for security upgrades.