What this is, and what it is not
This project exists because the systems described here are public record and effectively invisible anyway — scattered across agency spreadsheets, permit filings, and university archives. Putting them on one map changes what a resident can see about their own street. That is worth doing carefully, so here is exactly what this is.
| What this is | What it is not |
|---|---|
| A map of public infrastructure, institutions, contracts and policy. | A map of people. No layer here names, describes, or tracks any individual. |
| A historical and periodic record of where things have been observed. | A live tracker. Nothing here updates in real time, and nothing shows where anyone is. |
| Sourced from public records, with every source cited and every gap stated. | Complete. Every layer is partial, dated, or approximate — each one says how. |
| Informational tools for making records requests and contacting officials. | Legal advice. If you need legal help, talk to a lawyer. |
Transparency for systems, privacy for people
The whole project turns on one line: we point at power and infrastructure, never at people. Cameras, agencies, facilities, contracts, buildings, and policy are fair subjects for public scrutiny — they are funded publicly, they exercise public authority, and their records are public. Individuals are not.
This is a hard product boundary, not a preference. There is no field in our data schema that could hold a personal name, a case number, or an identifier for a private individual, and there will not be one. Where an upstream dataset mixes individual records into systemic data, the ingest takes the systemic part and drops the rest. Detainee-level data is permanently out of scope — it will never enter this repository, this map, or any export from it.
We do not surveil our own users
There are no accounts, no trackers, no third-party analytics, no advertising technology, and no cookies on this site. There is no server-side application at all: the whole thing is static files. When you use the "near me" lookup, the place list is downloaded to your browser and the search runs there — nothing you type is sent anywhere, because there is nothing to send it to and nowhere to log it. If you use browser geolocation, the coordinates are read into memory and discarded.
It would be incoherent to map surveillance infrastructure while quietly running surveillance infrastructure. So we don't.
About the camera layer specifically
The ALPR layer deserves its own warning. It is crowd-sourced from OpenStreetMap contributors, which means it is incomplete, historical, andprobabilistic. A camera may have been removed, moved, or re-aimed since someone mapped it. A street with no dot may still be covered.
Read it as evidence that a surveillance network exists and roughly where it has been observed — never as a guide to where you are or are not being recorded.
Errors and takedowns
If a record here is wrong, we want to fix it. Every layer links to its upstream source, so corrections usually belong upstream — at OpenStreetMap for cameras, or with the publishing agency for the rest — and will flow here on the next refresh. For anything else, open an issue on the repository.
Grassroots mapping of ALPR networks has drawn legal threats before; DeFlock received a cease-and-desist and refused it with EFF backing on First Amendment grounds. Our posture is to keep provenance clean, use only public data, publish the ingest code, and stay easy to mirror.
Licensing
The code is MIT. Data we compile ourselves is CC BY 4.0. Upstream layers keep their own terms, which are not ours to relax — the camera layer is ODbL, redlining is CC BY-NC-SA 4.0, and the federal records are public domain. The open data page lists which is which before you download anything.