Hackers breach Flock Safety camera, expose tracking of vehicles and people

Recovered data shows cameras detect pedestrians and bumper stickers, challenging company’s privacy promises

By LineZotpaper
Published
Read Time2 min
Hackers have pulled down a Flock Safety camera from a roadside, copied its data, and shared it with journalists, revealing that the company’s automatic license plate readers also detect people, bicycles, and even bumper stickers — casting new doubt on claims of on-device encryption and limited surveillance.

In a joint investigation with transparency nonprofit Distributed Denial of Secrets, 404 Media and WIRED obtained a near-complete copy of the storage from a Flock Safety camera after hackers removed the device from a roadway and extracted its contents. The hackers recovered an encryption key stored on the camera itself, allowing them to unlock videos of thousands of vehicle detections.

While the most sensitive parts of the camera’s storage remained encrypted and inaccessible, analysis of the recovered data reveals that Flock’s software explicitly detects people as well as vehicles, license plates, and bicycles. According to several weeks of recovered logs, the camera produced more than a million images from passing traffic. Its computer-vision system also isolated bumper stickers and graphics — in one case capturing an American flag patch on a motorcyclist’s saddlebag.

Flock Safety has marketed its systems as protected by on-device encryption. The breach exposes a flaw in that claim: the encryption key was stored on the device itself, making the data recoverable by anyone with physical access. The hackers told the outlets they plan to publish details of how they obtained the software, encouraging others to replicate the attack.

The findings offer an unprecedented inside look at how Flock cameras operate in practice, going far beyond the company’s public descriptions of its technology.

§

Analysis

Why This Matters

  • Flock cameras are deployed in thousands of communities across the US, often promoted as a crime-fighting tool; this breach reveals they collect far more data than simple license plate reads.
  • The recovery of an on-device encryption key undermines Flock’s security promises and raises questions about the safety of sensitive law enforcement data stored on these cameras.
  • If the hackers publish a method for others to copy the software, it could trigger a cascade of similar breaches, exposing tracking networks nationwide.

Background

Flock Safety is a private company that sells automated license plate reader (ALPR) cameras to police departments, homeowners’ associations, and businesses. The cameras are typically mounted on poles or traffic signals and are designed to scan passing vehicles. Flock has publicly stated that its systems use on-device encryption to protect data and that they only capture license plates — not people. The company has faced scrutiny from privacy advocates over the scale of its surveillance network, which logs tens of millions of vehicle detections daily. This hack provides the first independent verification of the camera’s actual capabilities.

Key Perspectives

Flock Safety: The company has maintained that its cameras are encrypted and limited to vehicle data. It has not yet commented publicly on this breach. Hackers/Transparency Groups: They argue that by exposing the camera’s true data collection, they are revealing the gap between Flock’s marketing and reality. They aim to make the exploitation method public. Privacy Advocates and Civil Liberties Groups: This breach confirms long-standing concerns that ALPR systems are expanding into general surveillance, capturing pedestrians and personal details without consent or oversight.

What to Watch

  • Whether Flock Safety issues a statement or software update addressing the on-device encryption key vulnerability.
  • How law enforcement agencies using Flock cameras respond — will they suspend use pending an investigation?
  • Whether the hackers follow through on publishing a full exploitation guide, and if others replicate the attack on cameras in other locations.

Sources

Zotpaper

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.