A Constitutional Proposal to Limit the Use of Flock Cameras: The Sought-Vehicle Rule

Watch the full episode here:

Automated license-plate readers present a deceptively simple civil-liberties problem.

Police officers have always been able to see license plates. They can run a plate they encounter on patrol, determine whether a vehicle has been reported stolen, check whether its registered owner has an outstanding warrant, or identify a vehicle connected to an investigation. Mike knows this firsthand from his years as a police officer. Running plates was routine police work.

But there is an enormous difference between a police officer observing a vehicle in public and the government constructing a searchable history of where millions of innocent people have traveled.

That distinction is at the center of the controversy surrounding Flock Safety cameras and other automated license-plate-reader systems.

The question is not whether technology should help police find stolen cars, locate missing children, or identify vehicles already connected to crimes. It should.

The question is whether accomplishing those legitimate objectives requires government to record everybody else.

It does not.

Automated license-plate readers can perform a task police officers have performed manually for decades, only much faster. Instead of an officer entering individual plate numbers into a computer, cameras can process enormous numbers of vehicles continuously.

That efficiency has obvious law-enforcement value.

It also changes the nature of the activity.

A police officer seeing your vehicle on Tuesday afternoon is one observation. A network of cameras recording your vehicle Tuesday afternoon, Wednesday morning, Thursday evening, and again the following week can become something very different: a history of your movements.

Once observations are retained and made searchable, government no longer merely knows that a particular vehicle passed a particular camera. It potentially has the ability to reconstruct patterns—where someone works, where someone worships, whom someone visits, what political meetings someone attends, what medical offices someone frequents, and where someone spends the night.

None of this requires the person being tracked to have done anything wrong.

That is the institutional problem.

The issue is therefore not really the camera. It is the database.

One common response to these concerns is to impose limits on how long license-plate data can be retained.

That is better than indefinite storage, but it does not address the underlying principle.

A government database containing seven days of innocent people’s movements is still a government database containing seven days of innocent people’s movements. Shortening the retention period reduces the size of the surveillance system. It does not change what the system is.

The more important question comes earlier:

Why should an identifiable government record of an innocent person’s location be created in the first place?

That question leads to a different approach—one Mike calls the Sought-Vehicle Rule.

The Sought-Vehicle Rule begins by separating two functions that are too easily conflated: using technology to find someone the government is already lawfully seeking, and using technology to collect information about everyone so that government might find it useful later.

Under Mike’s proposal, automated license-plate technology could still scan vehicles. But every observation would have to produce one of two results: YES or NO.

YES: The observed plate matches a vehicle already on an authorized sought list. That could include a stolen vehicle, a vehicle associated with an Amber Alert, or a vehicle already connected to a crime or legitimate investigation.

Only in that circumstance would the system be permitted to create an identifiable government record.

Once the system produces a legitimate match, the ordinary legal questions begin: how long the information may be retained, who may access it, whether a warrant is required for subsequent investigative use, whether it may be shared with other agencies, and under what circumstances it may become evidence.

But the predicate comes first. Government already has a lawful reason to be looking for that vehicle.

NO: The plate does not match a vehicle on the authorized sought list.

In that case, no identifiable government record would be created at all.

No stored plate number. No identifiable photograph. No location history. No timestamp attached to that vehicle. No vehicle description or other information capable of reconstructing the observation later.

The system sees the plate, determines that this is not a vehicle government is seeking, and moves on.

Nothing identifiable goes into a government database.

That is the critical distinction.

The Sought-Vehicle Rule is deliberately more protective than a conventional data-retention limit because it changes the sequence of government surveillance.

Under a broad collection model, government gathers information about everyone first and decides whose movements are interesting later.

Under the Sought-Vehicle Rule, government must have a legitimate reason for seeking a vehicle first. Only then does an identifiable observation become a government record.

In other words:

Suspicion should precede surveillance, not result from it.

This preserves the useful part of automated license-plate technology without requiring the creation of a mass database of innocent people’s movements.

Police can still use technology to find the proverbial needle. They simply do not get to build the haystack out of the daily lives of everyone else.

There is a tendency in debates over surveillance to say that because an individual observation is lawful, automating millions of those observations must also be lawful and unobjectionable.

That skips over the most important issue.

Scale matters.

A police officer can stand on a public street and see a license plate. A police department could theoretically assign officers to follow thousands of citizens around town, recording where they go every day. The fact that each individual observation takes place in public does not mean Americans should be indifferent to a government institution capable of assembling all those observations into searchable personal histories.

Technology has made collection dramatically cheaper, faster, and more comprehensive. That can make government more effective. It can also give government capabilities that previous generations never seriously had to confront.

Civil-liberties protections therefore cannot depend exclusively on asking whether government could theoretically have obtained a piece of information using older technology.

The more useful question is what institutional power we want government to possess.

Good safeguards should not depend on every future police chief, prosecutor, contractor, intelligence official, or elected politician exercising perfect judgment.

That is not a serious model of government.

A system that collects innocent people’s information and then promises not to misuse it is weaker than a system designed so that the information never exists.

That is the advantage of the Sought-Vehicle Rule.

It does not require us to choose between effective policing and civil liberties. It establishes a boundary between them.

Use the technology aggressively to find vehicles government has a lawful reason to seek.

For everybody else, do not create the record.

That is a relatively simple technological rule built around a much older constitutional instinct: government power should follow a legitimate public purpose rather than search indefinitely for one.

There is nothing inherently objectionable about technology making police work more efficient. A system that helps locate a stolen vehicle faster, recover a kidnapped child, or find a vehicle legitimately connected to a serious crime can produce enormous public benefit.

But technological capability is not its own justification for government authority.

The fact that government can cheaply collect information about millions of people does not establish that it should.

The proper response to increasingly powerful surveillance technology is not to prohibit every useful innovation. Nor is it to shrug and accept mass collection because the cameras happen to be pointed at public roads.

It is to establish the rule before the technology establishes the norm.

Find the people and vehicles government is lawfully seeking.

Leave everybody else alone.

That is not anti-police. It is not anti-technology. It is a guardrail designed to allow government to perform its legitimate functions without turning ordinary life into a permanently searchable government record.

And that is precisely where a free society should draw the line.

Watch the full episode here: