← Back to blog
Privacy April 11, 2026

We Aggregated 458,500+ Surveillance Records Across 183 Countries

We built Symvek Sentinel to answer a simple question: what’s watching you?

To answer it honestly, we did not collect anything ourselves. We integrated public-records sources covering 183 countries, normalized them into a single schema, and made the result searchable. No scraping. No proprietary data. Everything in Sentinel comes from public sources that other people, governments, and communities did the hard work of collecting.

The result: 458,500+ surveillance records aggregated worldwide. And we’re still expanding.

Where the data actually comes from

We do not aggregate cameras. We aggregate datasets. The credit belongs to the upstream sources, and getting that attribution right matters.

FLOCK community dataset (github.com/Ringmast4r/FLOCK) is our largest single source. The Ringmast4r/FLOCK project compiles ALPR camera locations and inter-agency data-sharing relationships from public records in the United States. It contributes the bulk of the US ALPR layer.

EFF Atlas of Surveillance contributes the US-jurisdiction layer: which law enforcement agencies use which surveillance technologies across 6,000+ US jurisdictions. ALPR, facial recognition, drones, ShotSpotter, body cameras, predictive policing, and more.

OpenStreetMap contributors are the global community-mapping layer. OSM volunteers have tagged hundreds of thousands of fixed surveillance cameras using man_made=surveillance, with particularly dense coverage across Europe, Asia-Pacific, and major metro areas.

DeFlock is the ALPR-specific community layer, complementing FLOCK with crowdsourced reports on automatic license plate readers.

20+ government open-data feeds contribute the high-confidence civic layer. TfL London (883 JamCams), Caltrans California (3,434 cameras across 12 districts), DriveBC (1,058 highway cameras), Ontario/Alberta/Quebec 511 networks, Toronto, Vancouver, Calgary, Leicester City Council, Austin TX, and many more. Government feeds are the highest-confidence source because each record is published by the operator.

What Symvek actually contributes

Symvek’s contribution is operational, not collection. We do four things on top of the upstream sources:

  1. Integration. One schema across dozens of feeds with different formats, coordinate systems, and update cadences.
  2. Deduplication. Removing records within 50 meters of each other across overlapping sources.
  3. Daily refresh. Government feeds and community datasets update on different cadences. We pull on a schedule and reconcile.
  4. Search. A single interface across the integrated dataset, with no tracking and no account required.

That is what Sentinel is. We did not map 458,500 cameras. We made the work of others queryable as one map.

Coverage by country (top 21)

CountryRecordsPrimary Types
United States290,826CCTV, ALPR, facial recognition
France60,983Speed cameras, CCTV
Spain10,080Speed cameras
Poland9,205Speed cameras, CCTV
Austria5,841Speed cameras
United Kingdom5,637CCTV, ANPR, speed cameras
Germany4,918Speed cameras, Autobahn CCTV
India4,084CCTV, speed cameras
Switzerland4,095Speed cameras
Sweden3,516Speed cameras
South Korea3,028CCTV, speed cameras
Canada2,930Highway CCTV, traffic cameras
Japan1,784CCTV, speed cameras
Australia1,548Speed cameras, CCTV
Belgium1,438Speed cameras
Czech Republic1,235Speed cameras
Portugal987Speed cameras
New Zealand895Speed cameras, CCTV
Netherlands658Speed cameras, CCTV
Norway398Speed cameras
Denmark21Speed cameras

The remaining records span 162 additional countries with thinner per-country coverage, drawn primarily from OpenStreetMap and FLOCK community contributions.

What the data shows

351,705 are CCTV cameras. Fixed, dome, and pan-tilt-zoom installations in public and private spaces.

50,500 are speed cameras. Automated enforcement across 20+ countries. France alone has more speed cameras than most countries have total surveillance points.

9,125 are ALPR readers. Automatic license plate readers that log every vehicle that passes. In the US, Flock Safety alone operates 80,000+ cameras across 49 states, scanning 20 billion plates per month. The UK has 1,226 ANPR cameras represented.

The rest includes facial recognition systems, gunshot detection microphones, body cameras, drones, cell-site simulators, and red light cameras.

What the data doesn’t show

This is what is in public records. The real number is much higher.

Clearview AI has indexed 70 billion faces from 3,000+ law enforcement agencies. Amazon Ring operates 20-30 million residential cameras. Retailers deploy facial recognition without disclosure. China has an estimated 600 million cameras that do not appear in any open dataset. None of these surface in public mapping databases.

Sentinel reflects what the FLOCK, EFF, OSM, DeFlock, and government communities can verify. The invisible surveillance infrastructure is orders of magnitude larger.

What you can do

Open the map and search your address. See how many records are within a kilometer. Toggle the heatmap to see surveillance density patterns. Pan across Europe to see France’s speed camera density. Zoom into London or Seoul.

The map runs entirely in your browser. We never know where you looked. No analytics, no cookies, no accounts. The watchers are already public. We just made them legible.

What comes next

We are expanding integration to additional government feeds as they come online and pulling new community sources as the FLOCK, DeFlock, and OSM communities expand them. Italy, Finland, Singapore, and more Australian states are in the pipeline. We are also building tools that perceive surveillance in your physical environment, not just the records that exist in public databases.

If you want to follow our progress, check back here or follow our work on the Sentinel page.

The cameras don’t go away. But now you can see what the public record already knows.