53.8% Location Precision: Swedish Wi-Fi System Spots Hidden Guns Without Cameras

Researchers at the Swedish Defense Research Agency have created an artificial intelligence system that can identify concealed handguns by analyzing how they distort Wi-Fi signals, according to Tom's Hardware. The system, named WiFireNet, operates without visible cameras or physical metal detectors, relying instead on the way wireless signals bounce off objects in a room.

The risk

The system offers a non-intrusive alternative to visible cameras and cumbersome metal detectors, according to Tom's Hardware. It identifies concealed handguns using Wi-Fi signals, which allows for detection without the need for direct visual observation or physical screening equipment. This capability enables the identification of specific objects based on their interaction with ambient wireless signals.

Who is exposed

Individuals in controlled indoor spaces where the WiFireNet system is deployed are subject to this monitoring. The prototype was tested in small indoor areas with up to four moving individuals. The system distinguishes between everyday items like phones and water bottles and concealed handguns, meaning objects carried by people in the monitored zone could be detected.

Cost and compliance

The research was detailed in a paper posted on arXiv on October 13, according to Tom's Hardware. The current prototype uses a modified implementation of the older 1999 IEEE 802.11a wireless standard. The researchers state that subsequent phases will utilize modern Wi-Fi 7 and Wi-Fi 8 networks to achieve higher-resolution channel-state information for improved detection capabilities.

What should you do?

Managers should assess whether existing Wi-Fi infrastructure could be repurposed for signal analysis without explicit authorization. Teams should review their physical security policies to ensure that non-intrusive detection methods align with privacy regulations and employee consent requirements.

The background in plain language

Wi-Fi signals do not just carry data; they physically interact with the environment. When these radio waves hit an object, they reflect, refract, and absorb in patterns unique to that object's shape and material. This data, known as channel-state information, creates a digital fingerprint of the space. By training AI on these patterns, systems can identify objects without seeing them, similar to how sonar maps the ocean floor using sound waves.

What to do and how to stay safe: WiFireNet

• Audit your network configurations to ensure channel-state information is not being collected or stored unnecessarily.
• Review privacy policies to address surveillance methods that do not involve visible cameras or explicit scanning devices.
• Monitor vendor updates for security features that prevent unauthorized analysis of wireless signal reflections.
• Educate staff on the limitations and capabilities of modern Wi-Fi networks regarding physical object detection.

Step-by-step guide: AI Governance Explained: Building Rules for Autonomous Systems

General security guidance from the Vector Tribune newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

How accurate is the WiFireNet system in detecting handguns?

The system achieved an average precision of 72.5% in controlled indoor tests, according to Tom's Hardware.

Does the system require cameras to work?

No, the system analyzes channel-state information from Wi-Fi signals and does not use cameras or metal detectors.

What happens to accuracy when locating the exact position of the weapon?

Precision drops to 53.8% under stricter localization requirements that pinpoint the weapon’s exact location on the carrier.