Flock Safety’s own patent indicates that its surveillance capabilities are more extensive than what the company has publicly claimed, challenging assertions that its cameras merely function as license plate readers.
The response to the Flock surveillance system has been rapid and, in many cases, quite effective. Activists and community members have pressured various states, counties, and law enforcement agencies to terminate their contracts with the company. As more regions are distancing themselves from Flock, the company has felt the need to advocate for its technology as a practical tool, asserting that it’s not as invasive as critics suggest. They’ve also mentioned the implementation of new “safeguards” aimed at preventing misuse by police officers.
However, Flock’s patent uncovers a system with capabilities that far exceed the company’s public disclosures.
In August 2022, the U.S. Patent and Trademark Office granted Flock a patent titled “System and method for object based query of video content captured by a dynamic surveillance network.” The patent credits Flock’s CEO, Garrett Langley, and co-founder Matt Feury as its inventors.
Flock’s Frequently Asked Questions state that the system is “not a live-tracking system” and that each captured image is merely a “standalone vehicle event.” Yet, the patent implies the opposite.
The patent describes Flock’s system as one designed to integrate various surveillance cameras—whether positioned at busy intersections, outside convenience stores, or within gated communities—into a unified, searchable, and “dynamic” network.
To quote from the patent, “the surveillance network is dynamic because embodiments of the system are not constrained by any geographic footprint associated with a set network of surveillance devices.” It allows different cameras to upload video feeds into the system.
This capability enables searches that extend beyond just license plates; it includes “face recognition data points.” The patent also mentions that the system can track and identify individuals based on attributes such as sex, race, clothing, height, and even weight:
For instance, if the identified object is a human, the detection module can analyze the image using a neural network designed to classify various aspects of individuals (male, female, race, etc.). Afterwards, the module might utilize other neural networks to further detail features like clothing type or physical measurements.
The patent outlines a scenario where a “subscriber” can identify and monitor a specific individual through an “alert triggering module.”
This system could potentially “recognize that a person seen in a video clip… bears resemblance to a subscriber’s teenage daughter and subsequently send an alert (such as a text message) to the subscriber’s phone.”
Contrary to what the patent states, Flock published a blog post in May, asserting that its cameras “do not use facial recognition” and are “not designed to identify people” in a general manner.
According to the blog, “They are built to identify vehicle information, not personal biometric data. They do not search for individuals, scan faces, or track people. This distinction is important because it clarifies the intended purpose of the technology. By focusing on vehicles instead of people, Flock enables investigations that are more straightforward and limited.”
In June, during an interview with Atlanta News First Investigates, Flock’s Chief Communications Officer, Josh Thomas, asserted that the Flock system does not track people.
“Tracking your complete movements is quite different from knowing you passed by one particular camera or a set number of cameras,” he stated. “Those are distinctly different.”
When the reporter pressed him on this point, Thomas reaffirmed his stance.
“We’re not tracking people,” he maintained.
