United States, Nationwide
US Secret Service
Licence plate reading from poor-quality images and video (forensic), supplied by Amped (as listed in the inventory)
OperationalOfficial publicationPAI-F65508F1
- Body
- US Secret Service
- Country
- United States
- Area
- Nationwide
- Technology
- Licence plate reading from poor-quality images and video (forensic)
- Supplier
- Amped (as listed in the inventory)
- Status
- Operational
In routine use by the named body, on the evidence of a public source. - Stage
- In use now
- Since
- January 2025
- Legal basis
- Not stated in the source
- Oversight
- Listed in the DHS 2025 AI Use Case Inventory (DHS-2626), where it is classed as not high-impact
- Evidence
- Official publication
Police force, government department or public body publication - Source
- Office of Management and Budget, 2025 federal AI use case inventory (DHS-2626)
- Last checked
- Record code
PAI-F65508F1
Notes
The DHS 2025 AI Use Case Inventory (entry DHS-2626, as consolidated by the Office of Management and Budget) lists a "License Plate Reader" use at the Secret Service as deployed from 7 January 2025, purchased from a vendor listed as Amped. The entry says it is used to read licence plate information from low-quality imagery affected by low resolution, poor lighting, motion blur or obstructions: image enhancement and optical character recognition highlight possible characters and combinations as a decision-support aid, and investigators generate and verify leads alongside their own analysis. It says the vendor trained a neural network on millions of synthetically generated and distorted licence plates, with no plate images scraped from the web, and that the vendor's validation was on Italian plates. DHS classes the use as not high-impact. Nextgov (15 October 2026) reported Secret Service chief information officer and chief AI officer Chris Kraft describing a deployed licence plate model that deciphers plates from poor-quality images or video. This is a forensic tool applied to images and video already held, not a network of plate-reading cameras.
Other systems used by US Secret Service
| Deployment | Country | Supplier | Since | Status | Checked |
|---|---|---|---|---|---|
| US Secret ServiceFacial recognition for investigative leads (Office of Investigations) | United StatesNationwide | NEC (algorithm used in DHS OBIM's IDENT system, per the DHS inventory) | — | Operational |
Licence plate reading from poor-quality images and video (forensic) elsewhere in United States
See them in the tracker| Deployment | Country | Supplier | Since | Status | Checked |
|---|---|---|---|---|---|
| ICE Homeland Security InvestigationsTRM Labs blockchain analysis | United StatesNationwide | TRM Labs | 2022 | Operational | |
| ICE Homeland Security InvestigationsCellebrite media classifier for digital evidence | United StatesNationwide | Cellebrite | 2023 | Operational | |
| US Immigration and Customs EnforcementAI mobile device forensics triage | United StatesNationwide | — | — | Proposed | |
| US Customs and Border ProtectionTRM Labs cryptocurrency analysis | United StatesNationwide | TRM Labs | 2025 | Operational | |
| US Customs and Border ProtectionAI triage of smartphone content during inspection | United States | — | — | Proposed | |
| Drug Enforcement AdministrationDrug Signature Program models (origin and manufacturing route of seized drugs) | United StatesNationwide | — | — | Operational | |
| Drug Enforcement AdministrationFriction ridge print comparison and print-card character recognition | United StatesNationwide | — | — | Operational | |
| Drug Enforcement AdministrationFinancial and cryptocurrency analysis tools (federal partner and commercial) | United StatesNationwide | — | — | Operational | |
| US Department of JusticeAI features in digital forensics tools (for example Cellebrite, Magnet Axiom, Griffeye) | United StatesNationwide | Cellebrite, Magnet Axiom and Griffeye (examples named in the inventory) | — | Operational |