Chapter 1, section iv · the Vigilant Solutions case
Hot List
In January 2016, Vigilant Solutions gave several Texas local governments license-plate readers and access to its database. In return the governments gave Vigilant their lists of outstanding warrants and unpaid court fines. A driver flagged by a camera on a police car could pay on the spot, plus a 25 percent fee to the company, or be arrested. The company kept every plate the police cameras read. This page runs a small version of that arrangement.
Synthetic. The town, the cars, the plates and every number on this page are made up. Nothing here describes a real place or person. The model's assumptions are labeled, and you can change them under Settings and assumptions.
- car
- plate on the hot list
- fixed camera
- patrol car with reader
- a read
Click or tap a car to see what the database has on it.
Pattern of life
—- Sightings
- 0 reads over 0 days
- Inferred home
- —
- Inferred daytime place
- —
- Check
- —
This is everything the database has on this car. No camera saw a driver, a face or a reason for a trip; it saw a plate at a time and a place. The home and daytime place are guesses from where the plate shows up overnight and on weekday afternoons. Paglen calls this a "pattern of life" signature.
Sightings by hour, most recent days
fixed camera patrol-car camera read while on the hot list
Latest sightings
Ledger
Stops per 100 cars
by the driver's home neighborhood
Arrests per 100 cars
driver could not pay on the spot
Paid at the roadside
fines plus fees, total dollars
Latest stops
Settings and assumptions
The system
Placed on Main St and 7th Ave first, then elsewhere.
Only patrol cars stop drivers. Fixed cameras just add reads.
The Texas contracts set this at 25 percent.
Assumptions not data
These numbers are invented to make the mechanism visible. Change them and see what moves.
0 means the same share everywhere. 100 means the share tracks household income steeply.
| Neighborhood | On hot list | Can pay on the spot |
|---|
Every assumption in the model
What to look at
- The cameras read every plate the same way. Most stored reads are of cars that owe nothing, and they go into the vendor's database anyway (see the share under the database count).
- Who gets stopped depends on the list, not the camera. The city writes the list. How the list is spread across neighborhoods is an assumption here; move the concentration slider to 0, skip ahead a few days, and watch the stop chart even out.
- Who gets arrested depends on who can pay on the spot. The same hit costs one driver a few hundred dollars and puts another under arrest.
- Set patrol cars to 0. The stops end, but the fixed cameras keep filling the database.
- Open a car's pattern of life. A few days of reads place most cars' homes within half a block. Night-shift cars get it wrong, and the database has no way to know.
What does this image do?
A plate read is a photograph no person looks at. It doesn't represent the car to anyone. It adds a row to a database, and when the plate string matches a line on the city's list, it puts a patrol car behind a driver and a card reader in front of them. The image's work is the stop, the payment, the fee and the arrest, plus every later sale of the stored row. That is Paglen's point in Chapter 1: "We no longer look at images—images look at us."
The chapter names two moves. First, individualization: every plate gets its own record of times and places. Second, reification: the record loses its ambiguity. A plate is on the list or it isn't, and the stop follows mechanically. Paglen's argument is that this objective-looking machinery serves the city's budget and the vendor's database. It does that at the expense of the residents least able to pay.
The book's conclusion shows where this kind of network went next. In 2025, a Texas sheriff's office searched more than 83,000 plate-reader cameras across the country to look for a woman suspected of self-managing an abortion.
Sources
- Trevor Paglen, How to See Like a Machine: Images After AI (Verso, 2026), ch. 1, "Invisible Images," section iv; and the Conclusion.
- Dave Maass, "'No Cost' License Plate Readers Are Turning Texas Police into Mobile Debt Collectors and Data Miners", Electronic Frontier Foundation, January 26, 2016. The source for the contracts with Kyle, Orange and Guadalupe County.
- "A Texas City Rescinds 'No Cost' License Plate Reader Deal For Being 'Big-Brotherish'", Electronic Frontier Foundation, February 25, 2016.
- Rindala Alajaji, "She Got an Abortion. So A Texas Cop Used 83,000 Cameras to Track Her Down", Electronic Frontier Foundation, May 30, 2025. Cited in the book's Conclusion.