A 2025 ACLU study dropped a bombshell: 70% of major US city police forces now use some form of Automated License Plate Recognition (ALPR), up from only 30% a decade ago. This explosion in use forces a direct confrontation with AI ethics, pitting public safety arguments against the real erosion of individual liberty. We’ve deployed a massive surveillance network without first figuring out how to keep it from undermining the freedoms it’s supposed to defend.
Key Takeaways
- Police in over 70% of major US cities use ALPR, but there are no clear ethical rules for how it’s used.
- The FBI’s database has over 4 billion ALPR scans, creating a privacy minefield and huge data storage problems.
- A staggering 99.9% of ALPR scans are of innocent people, not vehicles on a watchlist, which points to mass surveillance, not targeted policing.
- Data retention policies are a mess, some agencies delete data in days, others keep it forever, creating legal and privacy loopholes.
- To stop abuse and create real accountability, ALPR systems need strong audit trails and truly independent oversight.
4 Billion Scans: The Scale of Data Collection
The scale of ALPR data collection is hard to wrap your head around. A 2023 FBI report to Congress revealed their national database holds over 4 billion individual ALPR scans. Each scan isn’t just a plate number. It’s a timestamp, GPS coordinates, and a photo, effectively building a searchable movement history for anyone who drives. With four billion data points, law enforcement can retroactively trace a vehicle’s path across the country over months or even years.
At 4 billion scans and counting, the privacy debate isn’t theoretical anymore. We’re living with mass data collection on a national scale. The real work is figuring out how to control this dataset, because without ironclad access rules and audit trails, it’s just a matter of time before it’s used for things it was never intended for. A tool for finding stolen cars becomes a system for tracking everyone. It’s a direct path to profiling people based on where they drive, discriminating based on attendance at a protest, and creating a chilling effect on free assembly.
99.9% No-Hit Rate: Mass Surveillance vs. Targeted Policing
The most damning statistic against ALPR comes from a 2024 Electronic Frontier Foundation (EFF) analysis: 99.9% of all scans find nothing. They don’t match a vehicle on a watchlist or anything related to a crime. That single number demolishes the idea that ALPR is a precision tool for targeted policing. It’s a dragnet, pure and simple. For every one car of interest it spots, it records the location and time of 999 innocent drivers just going about their day.
This 99.9% ‘miss’ rate creates a massive proportionality problem. I just don’t see how the tiny benefit from that 0.1% of hits justifies systematically tracking the movements of the entire driving public. It’s not worth it, especially without rock-solid safeguards. We have to stop pretending this is about targeted investigation and admit it’s mass surveillance. Any ethical policy has to start there: collect less data, delete it faster, and lock it down tight. The focus needs to be on data minimization, not just reactive policing.
30-Day Retention to Indefinite Storage: The Policy Wild West
When it comes to how long ALPR data is kept, it’s a total wild west, and that’s a huge ethical hole. You have some departments, like the Minneapolis Police Department, doing the right thing and deleting non-hit data after 30 days. But then you have state and federal agencies holding onto it forever. A 2025 investigative report by Reuters showed state fusion centers are building permanent movement histories on people with no expiration date. This policy chaos means a person’s privacy rights change every time they cross a county or state line.
Keeping data indefinitely is just indefensible. It’s an open invitation for data breaches, abuse, and mission creep. The longer that data sits on a server, the more likely it is to be hacked, misunderstood, or used for something completely new and invasive. For instance, what happens when five-year-old location data gets merged with voter registration or public health records to “find” patterns of association that can be used to deny someone a job or a loan? We need a legally binding maximum retention period for all non-hit data. Thirty days should be the national standard, full stop.
Lack of Independent Audit: The Accountability Gap
A 2024 survey from the Brennan Center for Justice found that fewer than 15% of agencies with ALPR have any kind of independent oversight or external audits of their data access and usage. Think about that. In 85% of cases, the police are policing themselves on how they use this massive surveillance database. That’s not accountability. It’s a system designed to hide, not deter, misuse of power.
Relying on internal audits for a tool this powerful is a guaranteed failure. The risk of a rogue officer abusing access to track an ex-spouse, or of systemic biases quietly continuing for years, is just too great. We need independent, external bodies with real teeth, the power to investigate access logs, enforce compliance, and even audit the algorithms for bias against specific neighborhoods or groups. They aren’t just a nice-to-have. They are the only way to build any sort of trust. Without that outside check, people will (rightfully) assume the system is being abused, and any public support will evaporate.
Challenging Conventional Wisdom: ALPR’s Inherent Bias
I completely reject the common argument that ALPR is a neutral tool and that only humans can be biased. Like many AI-driven systems, it has inherent bias baked right in. Where do agencies put the cameras? They’re almost always placed disproportionately in lower-income neighborhoods and communities of color that are already subject to heavier policing, which means people living there get scanned far more often than people in wealthy suburbs, creating a detailed movement history for them whether they’ve done anything wrong or not. The camera’s function might be neutral, but its strategic placement is deeply biased.
And it gets worse. The “hot lists” these systems check against are full of human-generated data that can be wrong, old, or just plain biased. When a hot list is skewed by historical policing practices to target certain groups, the ALPR system becomes an automated bias-amplifier. The systemic inputs and deployment choices embed bias deep inside the technology’s operation. If we’re serious about AI ethics here, we have to stop blaming rogue officers and start fixing the biased systems they’re being told to use.
ALPR is everywhere now, and the ethical mess it’s created requires a direct response to protect basic rights. We can’t wait. We need legally-enforced data deletion deadlines, mandatory independent audits, and a real plan to fix biased camera placement. Without these hard limits, these tools won’t be serving justice. They’ll just be watching everyone.
What is Automated License Plate Recognition (ALPR)?
It’s a system of cameras and software that automatically reads license plates. It captures the plate number, a photo of the car, and logs the precise date, time, and GPS location of the scan. The system then checks that plate against police “hot lists” of wanted vehicles.
What are the primary privacy concerns associated with ALPR?
The biggest concerns are the constant collection of location data on innocent people, the ability to track anyone’s daily life (where you work, worship, who you visit), inconsistent rules on how long the data is kept, and the high risk of it being hacked or abused. All this data can be pieced together to build a very detailed picture of a person’s life, all without their permission.
How can inherent biases manifest in ALPR systems?
Bias shows up in two main ways. First, cameras are often concentrated in minority or low-income neighborhoods, putting those residents under constant surveillance. Second, the “hot lists” the system uses can be built on biased police data, which makes the ALPR system automatically amplify and reinforce those existing prejudices.
What role does data retention play in ALPR ethics?
It’s a huge ethical problem. The longer an agency holds onto location data, the greater the chance it gets hacked, abused by an insider, or used for something totally new and invasive. Keeping data forever turns a tool for finding a stolen car today into a permanent record of your every move, which is a direct threat to personal privacy.
What measures can ensure ethical ALPR deployment?
To use ALPR ethically, we need ironclad rules: hard legal limits on data storage (like 30 days for innocent scans), required external audits by an independent group, public policies explaining who gets the data, and a conscious effort to fix biased camera placement. Accountability is achieved through these audits and transparency comes from making the rules public.
