Allegheny County has more license plate readers than any other Pa. county. Here’s what’s at stake

Surveillance is a reality in the modern world. Security cameras record streets, stores, home entrances and even highways – and our phones keep track of where we go. Most of us likely don’t want to live like the characters in George Orwell’s novel “1984,” where an all-seeing government, Big Brother, is always watching. Yet technology used to fight crime has greatly increased the level of surveillance around the world.

In the United States, one technology in particular has become a focus of that debate: automated license plate readers.

These cameras, often mounted on poles at intersections and highway ramps, photograph every passing car. They record its plate, make, color and location, then store that information in searchable databases that police can check against lists of wanted vehicles. Flock Safety, one of the largest vendors, has cameras in thousands of U.S. communities, and their spread has prompted growing concern about privacy.

Allegheny County in southwestern Pennsylvania, home to Pittsburgh, has 827 automated license plate readers, more than any other county in the state.

Critics warn that the cameras can map people’s daily lives with little oversight, while police say they help find stolen cars and missing people. In the borough of Munhall, just outside of Pittsburgh, residents petitioned to remove two Flock cameras on Main Street.

Pennsylvania lawmakers from both parties have proposed at least nine bills, ranging from an outright ban of these devices to warrant requirements. In October the state House passed a bill that would limit their use and how long data is kept.

I study policing and public order in democracies, and I spent more than 30 years in police administration in India and with United Nations missions abroad. I’ve seen firsthand how new technologies can help police and why they need guardrails.

A spectrum of uses

How extensively police use surveillance technologies can fall anywhere across a spectrum. At one end, cameras can be positioned to monitor virtually every vehicle and person within a specified area. At the other, they are deployed sparingly and aimed only at specific vehicles or people, such as stolen cars or past offenders.

Mass surveillance may entail systemic collection of data of large numbers of people in the public space, including those not suspected of any violations. These can include networks of license plate readers that capture all vehicle traveling through an area, or cameras using facial recognition to analyze images of individuals in public spaces against existing databases.

Targeted uses include matching the vehicle numbers or faces of people of interest. These could be past offenders or people whose behavior raises suspicion, such as someone lingering near a bank late at night. Another example is designating certain places, such as airports, military bases, courthouses or power plants, as high-security areas and deploying surveillance technology there, with or without informing the public.

In the middle could be the sweet spot: calibrated use of such technology at an optimum level agreed on through laws or regulations. The level could be measured by the number of cameras in a city, the density of devices per square mile or even the terabytes of data collected.

Risks of mass surveillance

Mass surveillance carries the inherent risk of the state collecting extensive information on its entire population, or on large segments of it, which creates significant potential for misuse. The state could guarantee that the data would be handled responsibly. Even so, the data could be misused on purpose, or accessed by unauthorized persons because of vulnerabilities in data security.

There is also the fear of “surveillance creep.” That is when technologies adopted for one purpose expand into others. License plate readers, for example, are typically sold to communities as a way to find stolen cars, missing children and suspects in serious crimes. But police have also used them to monitor protests. According to records obtained by nonprofit advocacy group the Electronic Frontier Foundation, agencies across the country searched Flock’s network hundreds of times in connection with protests over a single year.

Potential benefits

The potential benefits of surveillance for crime prevention appear to be significant. A 2026 study of 216 law enforcement agencies found that after they installed fixed plate-reader networks, car thefts fell relative to other property crimes, and more theft cases were solved by arrest. In Atlantic City, New Jersey, a major expansion of plate readers didn’t reduce violent crime overall, but it was associated with fewer shootings, car thefts and property crimes.

Other law enforcement areas that could benefit from the technology include drug running and human trafficking.

In Pittsburgh, plate-reader data feeds into the police bureau’s real-time crime center. The data collected could be matched in real time with law enforcement and other databases to track the movement of offenders.

That could be used to prevent a wide range of crimes, from thefts and robberies to drug running and human trafficking, or to help solve cold cases.

A limited approach

Limited use of such technology has the benefit of curbing misuse while still using it in a controlled manner against known offenders and threats. That means using it on a person-to-person basis with prior legal or judicial approval or in high-security areas.

In this context, it would be most appropriate to take three steps to address law enforcement data collection generally:

• Keep requiring police to get court warrants backed by probable cause for cellphone location data – records showing which cell towers a phone has connected to. That includes closing the loophole that lets agencies buy the same data from commercial brokers without a warrant.
• Put in place federal or state laws restricting the use of facial recognition, with suitable judicial oversight.
• Have county and municipal agencies regulate the use of surveillance cameras in public areas.

Pennsylvania’s proposals span this range and beyond. State Sen. Dan Laughlin, an Erie Republican, would ban the cameras outright. U.S. Rep. Pat Harrigan proposes legislation requiring police to get a warrant before searching plate-reader data. And the bill the state House passed in October sits in the middle, allowing the cameras but limiting their uses and how long data is kept.

Guardrails

Decisions regarding the use of mass cameras should not be left to law enforcement agencies. They are interested in crime control, which could color their judgment. Those decisions are best left to elected representatives at the federal, state and local levels, with input from legal experts and relevant law enforcement agencies.

In my view, mandatory public consultation before deploying such systems in a community would improve the legitimacy of the program and of the implementing government. Regular audits are a common safeguard when government agencies use intrusive technology. Pennsylvania’s House-passed plate-reader bill, for instance, would require agencies to adopt audit policies. Such audits could happen as often as twice a year, with information and analyses available to the community or to affected individuals on request.

Transparency of the process would go a long way toward enhancing public trust. The opposite can erode it. At Penn State, for example, the university initially declined to say how many plate-reader cameras it had installed or where, drawing criticism from students before it ended its contract with Flock.

Another precaution is to keep the data from being available in raw form to nongovernmental entities, especially corporations. Location records can be valuable to businesses. Companies already use plate-reader data for purposes such as verifying a loan applicant’s home address.

As Pennsylvania lawmakers weigh new limits, the choices they make may help determine how close any of us come to living in an Orwellian state.

Read more of our stories about Pittsburgh and Pennsylvania.