A sheriff's investigator was accused of running roughly 3,000 Flock searches on his ex-girlfriend's vehicle.
Authorized law-enforcement account → thousands of alleged personal searches → criminal investigation and arrest
EVIDENCE LIBRARY · GOVERNMENT
What happens when public systems generate a lead or profile?
Documented cases involving police, public agencies, courts, public databases, and government-accessible surveillance.
Chicago-area departments investigated officers for allegedly using Flock searches for personal purposes.
Police ALPR access → alleged searches involving former partners / employer → internal investigations
Wilmer Catalan-Ramirez: an unverified gang designation reached immigration enforcement.
Designation challenged → CPD said it could not verify it
50,176 dispatched alerts; documented gun-crime evidence appeared in a small minority of responses.
9.1% of dispositions documented gun-related criminal evidence
The Strategic Subject List assigned people risk scores for becoming a victim or offender in a shooting.
OIG: unreliable scores and tiers among major concerns
Robert Williams: a false facial-recognition lead became a wrongful arrest.
Wrongfully arrested → case settled → Detroit policy changed
Porcha Woodruff: Detroit faced another wrongful-arrest allegation after a facial-recognition match.
Eight months pregnant when arrested · civil litigation followed
A plate-reader misread reportedly put an innocent couple at gunpoint with their infant in the vehicle.
Wrong plate read → high-risk stop → error discovered
Michael Williams spent 11 months in jail while ShotSpotter evidence helped support a murder case that prosecutors ultimately dropped.
11 months jailed → case dismissed for insufficient evidence
Chicago's gang data included contradictory, incomplete and implausible records—with no ordinary way for a person to know or challenge a designation.
Bad data + broad access + weak correction process
Michael Oliver became another early Detroit wrongful-arrest case after a false facial-recognition match.
False match → wrongful arrest
Nijeer Parks spent ten days in jail while physical evidence that could have excluded him remained available to investigators.
False face match → 10 days jailed → civil challenge
By 2026, the known wrongful-arrest problem had grown beyond a handful of Detroit cases.
14 known wrongful arrests catalogued by ACLU in May 2026
Wrong reads, wrong hotlist entries and stale records have repeatedly sent police toward innocent motorists.
Different failure source → same roadside consequence
Illinois restricted how ALPR data could be used. A state audit later found that CBP had nevertheless gained access to Illinois Flock data.
State restriction → prohibited access found → access ordered terminated
Illinois ALPR access became entangled with out-of-state abortion and immigration investigations.
Interstate network access → Illinois legal safeguards tested
Oak Park's own published oversight data exposed erroneous Flock-prompted stops.
Local oversight made the error rate visible
A character-recognition error reportedly turned a family trip into a guns-out police stop.
O / 0 confusion → wrong vehicle alert → high-risk stop
A reported 7-versus-2 plate error escalated beyond a routine traffic stop.
Wrong character → detention at gunpoint → police dog deployed
An ALPR investigation reportedly contributed to an innocent driver spending 13 days in jail.
Vehicle association → arrest → 13 days jailed → error discovered
Denise Green shows that the high-risk false-hit problem predates Flock.
False stolen-car alert → guns drawn → handcuffed and detained
An officer used law-enforcement plate-reader access to track his estranged wife.
Authorized access → personal surveillance → criminal consequences
Illinois required warrants. CPD's Stingray unit nevertheless kept no deployment log—and a city-retained attorney advised against creating one.
Legal safeguard → no usable audit trail → compliance difficult to verify
Police used a cell-site simulator to find Kerron Andrews inside a home. His defense initially wasn't told how police had located him.
Secret location technique → discovery fight → evidence suppressed
Tallahassee police used Stingrays roughly 200 times before 2010 without disclosing Stingray use to a judge to obtain a warrant.
~200 deployments → warrantless / undisclosed technique → secrecy challenged
Sarasota acknowledged Stingray records existed. Before they could be inspected, the U.S. Marshals Service took possession of them.
Records requested → inspection arranged → federal agents take records → litigation follows
Records later identified 1,835 Florida state cell-site-simulator uses and hundreds of Tallahassee investigations.
Secretive technology ≠ rare technology
CPD bought tools for large-scale social-media monitoring and used social media to monitor political protests and gatherings.
Public posts → automated monitoring → political activity enters police intelligence workflow
After 2020 protests, Chicago created a 20-person unit for round-the-clock social-media monitoring—and disclosure of how it worked required another lawsuit.
Program publicly announced → records withheld → litigation and contempt order → policy disclosed