There is a coffee shop on nearly every American corner, and almost all of them offer the same thing alongside the espresso: free WiFi. Most of us connect without thinking twice. Some of us don’t even bother connecting we just sit there, phone in a pocket or turned off entirely, nursing a latte.
Here’s what the router at the next table may already know about you, whether you’ve touched your phone or not.
In 2026, researchers at the Karlsruhe Institute of Technology (KIT) in Germany announced findings that quietly reframe one of the most familiar pieces of technology in modern life., researchers at the Karlsruhe Institute of Technology (KIT) in Germany announced findings that quietly reframe one of the most familiar pieces of technology in modern life.
Using standard WiFi signals combined with artificial intelligence, they identified individuals with very high accuracy reportedly above 99%, not by tracking their devices, not by reading their browsing history, but by reading their bodies., not by tracking their devices, not by reading their browsing history, but by reading their bodies. The human body, it turns out, bends radio waves in a way that is as individual as a fingerprint. Walk into a room with a WiFi router running, and you leave a signature behind. Every time.
The findings are scheduled to be presented at an upcoming ACM Conference on Computer and Communications Security (CCS), and the researchers are calling for updates to WiFi standards to limit how this capability can be exploited., and the researchers are calling for updates to WiFi standards to limit how this capability can be exploited.
What Beamforming Feedback Information Actually Is
Most people have never heard of beamforming feedback information, which is probably fine, it’s a technical detail that lives quietly inside the WiFi 5 standard that most routers sold in the last decade already use. Beamforming is the process by which a router directs its signal more precisely toward connected devices rather than broadcasting in every direction at once. It’s an efficiency feature, something that makes your video call smoother and your streaming less likely to buffer.
To do this, the router and your device exchange a small, routine packet of data, the beamforming feedback information, or BFI, that describes the radio environment between them. Think of it as the router asking “what does the signal look like from where you are?” and your device answering. This exchange happens constantly, automatically, as a normal part of how modern WiFi works.
The KIT researchers found that BFI captures something more than network conditions. It captures the physical shape of the space, including the people in it. Human bodies absorb and scatter radio waves in ways that depend on their size, posture, and movement. That pattern is distinct enough, and consistent enough, that an AI trained on it can learn to tell one person from another without ever seeing their face, reading their name, or touching their device.
The Part That Changes Everything
And here’s the thing that stops you in your tracks: you don’t have to be connected to anything. You don’t have to have your phone on. You don’t even have to have a phone. Your body alone, moving through a space where a WiFi router is operating, produces a radio image that the KIT system identified individuals correctly at a rate the researchers described as exceeding 99%..
That’s not a prototype result from a controlled lab with identical subjects walking in straight lines. That’s the main experimental finding. The precision is high enough that it crosses from “interesting research curiosity” to “deployable surveillance infrastructure”, and the infrastructure is already everywhere.
Every airport gate. Every chain restaurant. Every office building lobby. Every library reading room. Every hotel corridor. Most of them have been running WiFi 5 or later for years. The routers are already there. The BFI signals are already being generated. The only thing missing, until now, was software trained to read them as human identifiers.
Researchers are quick to note that this capability is not yet built into commercial systems, and that deploying it would require intentional effort and access to the router’s data stream. But the point of the KIT study is precisely that the raw material, the signal, already exists in every environment where WiFi operates. There is no opt-out for the radio waves.
What This Means in Ordinary Life
We have spent years managing our digital privacy through familiar controls. Turn off location tracking. Use a VPN. Connect to guest networks. Decline to share your name or email when a café asks for a login. Those tools address the data your devices voluntarily transmit.
None of them address what your body transmits just by being present.
The KIT researchers are calling for WiFi standards bodies to consider modifications that would limit how BFI data can be accessed or stored, a sensible starting point, though standards revisions tend to move slowly while technology moves fast. The concern is not that your neighborhood coffee shop is currently running an AI identification system. The concern is that the building blocks for one are already in place, and they have been for years, and most people had no idea.
We gave WiFi a pass because it seemed passive. It’s the thing that lets you watch a movie on the couch, not the thing that watches you back. That assumption, it turns out, was only half right.
How strange it is to realize that the same signal keeping your laptop connected has been quietly reading the room, and everyone in it, all along.
The KIT team’s work is a warning and a prompt, not a reason to panic. But it’s worth sitting with for a moment the next time you settle into a chair at a café, phone face-down on the table, convinced you’re off the grid.
The router doesn’t need your phone. It already has you.
Source
This article was researched, written, and edited by our human editorial team. AI tools were used in a limited research-assistant capacity. All claims were independently verified.
