EN

Face recognition in Russia: cameras in the metro, "Bezopasny Gorod (Безопасный город)"

· · The Ateo Digital editorial team

Facial recognition in Russia is deployed primarily in Moscow: according to various estimates, the city has about 220,000 street and building-entrance cameras connected to a video surveillance system with recognition. The metro uses Face Pay — a fare payment system based on facial recognition, connected to a biometric database. In the regions, the Bezopasny Gorod (Безопасный город) program operates, combining municipal and private cameras on one platform.

Use is divided into two modes: real time (searching for wanted people, operational activities) and after the fact (analyzing recordings from events on previous days). The second mode is especially visible after mass rallies — detentions based on camera recordings have taken place in Moscow just a few days after protests.

Below is a neutral overview of how the system is technically organized, what is included in facial databases, and what practical options a user has to opt out.

In brief

  • Moscow has about 220,000 cameras with connected recognition; in the regions, there is the Bezopasny Gorod program.
  • The metro uses Face Pay — fare payment by face through biometrics.
  • The facial database is formed from the EBS, passport photos, and open social networks.
  • Use: real time (searching for wanted people) and after the fact (analysis of recordings).
  • You can refuse Face Pay or disable it through the Moscow Metro app.

How the facial recognition system is deployed

The system is organized differently in different parts of the country, but the basic principle is the same: cameras transmit a video stream to centralized processing, where an algorithm searches for matches against a database.

Moscow: street and building-entrance cameras

The city video surveillance system combines cameras on streets, in entrances of residential buildings, on public transport, and at housing and utilities facilities. The stream is processed centrally, and recognition is connected as a separate module. The Ministry of Internal Affairs and operational services have access to the system; requests are processed under a simplified procedure.

Metro and Face Pay

The Moscow Metro has turnstiles that support payment by face. To connect, you need to upload a photo and link a bank card in the Moscow Metro app. After activation, passage through a turnstile is recorded as “person by face X passed at Y.” Cameras in train cars and at stations operate in parallel.

Regions: Bezopasny Gorod

The program combines municipal cameras, cameras on transport, and cameras in schools and hospitals. Coverage and recognition quality depend heavily on the region. Large cities (Saint Petersburg, Kazan, Yekaterinburg, Novosibirsk) have developed infrastructure; in small cities, recognition often does not work or works only at a limited number of points.

Sources of the facial database

The reference database is formed from several sources: the EBS (Unified Biometric System), passport photos from Ministry of Internal Affairs databases, and photos from open social networks (based on mass scraping results). Recognition accuracy on good frontal images is around 99%; from side angles, in poor lighting, or with a mask, it is significantly lower.

Where and how cameras with recognition are used

ContextWhat is recordedWhat can be done
Street camera in MoscowPassing through the frame, time, locationRoute outside camera zones; a hat/mask reduces accuracy
Building-entrance cameraEntry/exit from the building, timeBasic recording; a separate request to the management company to anonymize data
Metro without Face PayPassage through station cameras (without linking to a card)Do not connect to Face Pay; monitor the list of permissions in the app
Metro with Face PayPassage linked to a card and accountDisable Face Pay in the Moscow Metro app
Airport, railway stationCameras with expanded access rights for security servicesMinimize public data about your route
Mass rallyVideo recording with after-the-fact recognitionConsider the risks when participating; do not post on social networks

How to check right now

It is impossible to directly check whether you are “connected to the recognition system” — the user has no interface to the city database. But you can check two controllable channels: Face Pay in the Moscow Metro app (the payment section shows whether biometric passage is activated) and your personal account on Gosuslugi (Госуслуги) in the biometrics section — the status of submitted biometrics in the EBS.

At the same time, it is worth checking yourself against risk registries (the Ministry of Internal Affairs wanted database, the list of terrorists and extremists): inclusion in one of them sharply increases the probability that city cameras will process your face in the background.

Check full name in the Ministry of Internal Affairs wanted registry

Warning signs vs false alarm

A warning sign is appearing in the wanted database or involvement in public cases. In this case, city cameras with recognition work as a passive “trigger”: when a face appears in the frame, the system sends a notification to an operational service. Minimizing public movements and getting legal support are the standard set of actions.

A false alarm is a general feeling that “I am being filmed everywhere.” There are indeed many cameras, but targeted recognition is enabled on specific grounds. For most people, this is background infrastructure without directed attention.

What to do next

  • If you do not use Face Pay, do not connect it and do not upload a photo to the Moscow Metro app.
  • Already connected to Face Pay? Disable it in the app and submit a request to delete biometric data.
  • Check the biometrics section on Gosuslugi — whether there is an uploaded template and whether you can submit a refusal.
  • Minimize the number of frontal photos in open social networks — this lowers the quality of the reference image for scraping.
  • Keep in mind that refusing Face Pay does not disable street cameras — it only reduces the “face → bank card” link.
  • At elevated risk levels, consult a lawyer before planning routes and trips.

Conclusion

Facial recognition in Russia is real infrastructure deployed at scale, especially in Moscow. It is technically impossible to fully “get away from cameras” in an urban environment, but controllable channels (Face Pay, EBS) can be disabled and risks can be assessed. For people outside public cases and registries, this infrastructure works as background; for people in registries, it is an active search channel, and the strategy here is determined by a lawyer, not technical life hacks.

Check through Wanted Radar