A look at ~123 police-run US real-time crime centers, which use CCTV, facial recognition, and social media monitoring, as critics warn of creeping surveillance
Context & Ripple Effects
The Wired census of ~123 police-run real-time crime centers is the latest checkpoint in an arc that began when The Intercept documented [[a:950007|AI-enabled video surveillance networks letting private businesses and homes pipe camera feeds straight to police headquarters]]. What was then a scattered practice has since industrialized: Fusus now links a town's cameras into a single AI-analyzed hub across 2,400 locations, showing how the center model scales through vendors rather than city-by-city procurement.
The accountability side of the ledger is also established ground — New Orleans police ran [[a:885931|200+ facial recognition cameras for two years in apparent violation of a 2022 city ordinance]], and Engadget's investigation into 100K+ Flock license plate readers found security flaws and police misuse. Against that record, the new survey's finding that these centers fuse CCTV, facial recognition, and social media monitoring lands less as novelty than as confirmation that oversight keeps trailing deployment.
First-order effects
- Residents in the ~123 cities hosting these centers are monitored through fused CCTV, facial recognition, and social media feeds run from police headquarters, while the departments operating them gain a centralized investigative capability built substantially on privately owned cameras.
Second-order effects
- Surveillance vendors get a repeatable municipal sales motion — Fusus's 2,400-location footprint shows the center model spreading through third-party hubs rather than bespoke police builds, pressuring rival suppliers to bundle camera integration and AI analysis the same way.
- Cities that passed restrictions, like New Orleans's 2022 ordinance, face an enforcement problem: the documented pattern of police running banned facial recognition scans anyway forces local lawmakers to decide whether ordinances need audits and penalties or remain advisory.
Third-order effects
- If the pattern holds, US urban policing consolidates around vendor-operated fusion hubs where the effective surveillance perimeter is set by camera owners opting in — not by public vote — pushing regulation toward governing data flows and vendor contracts rather than individual tools like cameras or plate readers.
The trend: US policing is consolidating fragmented private and municipal cameras into AI-fused real-time crime centers faster than local ordinances can constrain them.