When “Public Safety” Becomes a Data Pipeline: Why Flock Safety’s Contract Language Should Alarm Every American
- BCS Resident
- May 26
- 5 min read
Editor’s Note
This article is based on a review of the actual contract agreement between the City of College Station and Flock Safety regarding the deployment of Flock Safety surveillance infrastructure and services.
For transparency, the full contract PDF referenced throughout this article should be included alongside this post so readers can review the original language themselves, including the clauses related to:
anonymized data,
machine learning,
data retention,
third-party sharing,
and future “Flock offerings.”
The analysis in this article is based directly on the language contained in that agreement as well as publicly reported controversies involving Flock Safety’s disclosures and data-sharing practices.
Across the country, cities are rapidly installing automated license plate reader systems marketed as tools to fight crime, recover stolen vehicles, and assist police investigations.
One of the largest companies in that space is Flock Safety.
At first glance, the pitch sounds simple:
cameras capture license plates,
police receive alerts,
crimes get solved faster.
And to be fair, these systems have helped recover vehicles and identify suspects in serious criminal cases.
But buried inside many of these agreements is contract language that goes far beyond simple crime prevention.
After reviewing the City of College Station’s contract with Flock Safety, one clause stood out immediately:
“Flock shall have the right to collect, analyze, and anonymize Customer Data” for purposes including “training of machine learning algorithms,” improving services, and creating “other Flock offerings.”
That clause should concern every citizen — not because technology itself is inherently evil, but because of what this kind of language enables when combined with nationwide surveillance infrastructure, AI analytics, and weak public oversight.
The Public Thinks These Cameras Just Read License Plates
That is not what the contract actually describes.
The agreement defines Flock Services as including:
automatic license plate detection,
alerts,
audio detection,
searchable image records,
video,
and sharing of “Footage.”
The system also references “Vehicle Fingerprint” technology, which goes beyond simply reading license plates.
Modern vehicle fingerprinting can include:
make and model,
color,
roof racks,
bumper stickers,
decals,
visible damage,
wheel styles,
and other identifying characteristics.
In practice, this allows systems to track vehicles even when:
plates are obscured,
switched,
unreadable,
or absent altogether.
This is no longer merely a camera system. It is behavioral infrastructure.
The Most Important Distinction: Raw Data vs. Derived Intelligence
Flock repeatedly states in its contracts that it does not own or sell raw customer data.
But that reassurance hides a much more important detail.
The agreement grants Flock:
perpetual rights,
worldwide rights,
royalty-free rights,
to use anonymized and derived data indefinitely.
That means the company can likely:
run large-scale analytics,
train AI systems,
develop predictive models,
generate statistical intelligence,
and create future commercial products from the collected data.
And the contract does not appear to require public disclosure of:
what studies are conducted,
what conclusions are drawn,
what models are trained,
or what future systems emerge from that analysis.
The raw footage may belong to the city.
But the intelligence extracted from the system may effectively become part of Flock’s long-term proprietary ecosystem.
“Anonymized Data” Is Not the Safety Guarantee People Think It Is
Tech companies frequently use the word “anonymized” to calm privacy concerns.
But mobility data is notoriously difficult to truly anonymize.
Repeated travel patterns can reveal:
where someone lives,
where they work,
where they worship,
medical visits,
political activity,
social relationships,
and daily routines.
Even if names are removed, pattern-of-life analysis can often re-identify individuals using enough location data points.
This is a well-known problem in modern data science and surveillance research.
And unlike a single parking-lot security camera, Flock operates a massive interconnected network spanning thousands of communities nationwide.
That changes the scale of what becomes possible.
Why Trust Matters More Than Ever
Some people may still argue: “If the company acts responsibly, this isn’t a problem.”
But that argument depends entirely on public trust. And recent controversies involving Flock Safety suggest that trust may not be as secure as many communities assumed.
According to reporting from 404 Media, Customs and Border Protection had access to more than 80,000 Flock-connected cameras nationwide through pilot programs and data-sharing arrangements.
Another 404 Media investigation reported that a federal DEA agent allegedly used a local officer’s credentials to conduct immigration-related surveillance searches without that officer’s knowledge.
Whether one supports or opposes immigration enforcement is beside the point.
The issue is transparency.
Many communities believed they were approving a local crime-fighting tool — not participation in a broader federal surveillance ecosystem.
Oshkosh, Wisconsin: A Warning Sign for Every City
The trust issue became even more serious in April 2026.
The Oshkosh Common Council rescinded its Flock contract less than 24 hours after approving it following what city officials described as false or misleading statements about the system’s capabilities.
According to reporting from WBAY News, Flock representatives told council members the system could not create heat maps tracking vehicle movement.
The next day, Oshkosh Police Chief Dean Smith publicly stated that city staff verified the system did create heat maps and that the information presented to the council was inconsistent with reality.
One council member stated:
“I don’t like being lied to.”
Another apologized publicly for supporting the contract based on incomplete information.
The council then voted unanimously to rescind the agreement.
Flock later responded that the issue was a misunderstanding involving terminology and argued that the system does not create “pattern of life” tracking in the legal sense.
But the damage was already done.
The core issue was no longer just surveillance technology.
It was whether public officials were receiving complete and accurate information before approving expansive data systems.
Surveillance Systems Rarely Stay Limited
History shows that surveillance technologies almost always expand beyond their original purpose.
Tools introduced for:
anti-terrorism,
child protection,
stolen vehicle recovery,
or violent crime investigations,
often evolve into broader systems of monitoring and behavioral analysis.
The danger is not necessarily what the system does today.
The danger is what becomes possible:
after years of data collection,
after AI systems mature,
after databases merge,
and after public oversight falls behind the technology.
And once machine learning systems are trained on years of movement data, the intelligence derived from that data cannot realistically be “unlearned.”
The Real Problem: Oversight Has Not Kept Pace With Technology
Most city councils are not staffed by:
AI researchers,
surveillance experts,
privacy engineers,
or large-scale data scientists.
Yet local governments are increasingly approving contracts that may contribute to:
nationwide movement intelligence systems,
behavioral analytics platforms,
AI training pipelines,
and future predictive technologies.
Meanwhile, ordinary citizens often have:
little technical visibility,
limited audit rights,
no access to derived analytics,
and minimal understanding of how these systems actually evolve over time.
That imbalance should concern everyone — regardless of politics.
Read the Contract Yourself
One of the most important things citizens can do is review the actual agreements their local governments are signing.
The clauses discussed in this article are not speculative. They come directly from the City of College Station’s contract with Flock Safety, including language related to:
anonymized data,
machine learning,
third-party sharing,
retention periods,
and future product development.
Readers are encouraged to examine the full contract themselves rather than relying solely on summaries or marketing materials.
This Debate Is Bigger Than Flock Safety
This issue is not ultimately about one company.
It is about the future relationship between:
citizens,
governments,
corporations,
AI systems,
and mass surveillance infrastructure.
Society is rapidly building systems capable of recording, analyzing, and modeling human movement at unprecedented scale.
The question is no longer: “Can these cameras help solve crimes?”
The real question is:
“What safeguards exist when companies are allowed to derive long-term intelligence from massive behavioral datasets while operating largely outside public visibility?”
Right now, the answer appears increasingly unclear.
Comments