DRAFT ONLY — SUBJECT TO REVISION
Sarah T. Hamid
Director of Strategic Campaigns
Electronic Frontier Foundation
Assessing the
Effectiveness, Limitations,
and Impacts of Continuous
Vehicle Monitoring
Tuesday, October 6th, 2026 | Richmond, CA
DRAFT ONLY — SUBJECT TO REVISION
15 Years of Research on
Automated License Plate
Readers
Since 2012, EFF has pioneered the
documentation of Automated License Plate
Reader systems through public records
requests, technical analysis, and the Atlas of
Surveillance—a searchable database
documenting over 15,000 police surveillance
technology systems across the United
States.
DRAFT ONLY — SUBJECT TO REVISION
Understanding ALPR System Architecture
5-Step Automated Process:
1. Image Capture: Cameras photograph
license plates as vehicles pass.
2. Image Processing: Software isolates
plates and reads characters (OCR).
3. Data Transmission: Plate data is sent to
central servers.
4. Database Comparison: Plate data is
checked against relevant databases.
5. Storage & Analysis: All info is stored for
future use, reporting, and access
through the user portal.
DRAFT ONLY — SUBJECT TO REVISION
ALPR Deployment Types
Fixed: Mounted for
constant monitoring.
Mobile: On vehicles for
scanning while driving.
Portable: Quick-deploy
units for temporary use.
All use advanced cameras,
often with infrared, to
capture clear images day
or night.
DRAFT ONLY — SUBJECT TO REVISION
Emergence + Growth of ALPR Technology
1970s-1980s: Early Development and First Commercial Contracts The first systems were expensive, bulky, and difficult to use. They required specific lighting and camera angles, limiting their effectiveness and widespread adoption. 1990s: Market Expansion and Cost Reduction
ALPR arrived in the United States in 1998 via U.S. Customs and Border Patrol for border security.
2000s-Present: Industry Consolidation and New Market Entrants As computing power became cheaper and more efficient, the price point of ALPR devices decreased, and ALPRs were marketed for a broader range of use cases, not just policing.
Al-Enhancements Era: Cloud-First and Specialized Solutions The integration of AI paved the way for feature updates like "vehicle fingerprinting," a method of identifying a car using more than just its license plate.
DRAFT ONLY — SUBJECT TO REVISION
Flock
Safety: A
Surveillance
Social
Network
DRAFT ONLY — SUBJECT TO REVISION
Data Sharing Network Layers
DRAFT ONLY — SUBJECT TO REVISION
Pathways to Unauthorized Access
● Formal pathways: Unauthorized agencies can get access through established systems, contracts, or legal process. ● Informal pathways: Unauthorized agencies can rely on
local partners, side-door searches, or unofficial
arrangements.
● Unauthorized pathways: Access can happen through
misconfiguration, compromised accounts, or vendor
actions outside local visibility.
DRAFT ONLY — SUBJECT TO REVISION
Understanding Effectiveness
● Do arrests, recoveries, and case assists overlap?
● How many arrests produced sustained charges or convictions? ● How often was ALPR information indispensable rather than merely available?
● How many scans, searches, and alerts produced no useful result? ● Did clearance or recovery rates improve over a valid pre-ALPR baseline? ● How did Richmond compare with regional crime trends?
● What was the full cost per verified outcome?
DRAFT ONLY — SUBJECT TO REVISION
Structural Limitations of ALPR Leads
An ALPR alert is an investigative lead—not proof, identity, or probable cause by itself.
● A correct plate read does not establish who was driving. ● Vehicles may be shared, rented, borrowed, recently sold, or using altered plates.
● Alerts may reflect camera misreads, stale source lists, or incorrect manual entries.
● More alerts do not produce more safety without verification, staffing, and follow-up.
● An unverified mismatch can turn a database error into an armed police stop.
DRAFT ONLY — SUBJECT TO REVISION
The Impact Begins at Collection
● ALPRs investigate everyone first: the system records vehicles because they pass a camera, not because their occupants are suspected of wrongdoing.
● Repeated observations can reveal homes, workplaces, healthcare visits, religious practice, political activity, and personal relationships. ● Historical search converts routine movement into information available for later investigation.
● Errors and misuse carry greater consequences for communities already subject to disproportionate police contact.
DRAFT ONLY — SUBJECT TO REVISION
DRAFT ONLY — SUBJECT TO REVISION
“Given how pervasive,
sprawling, and
ungovernable ALPR
sharing networks have
become, the only feature
update we can truly rely
on to protect people's
rights and safety is no
network at all.”