Responsible Innovation
Vehicle Intelligence
A camera-based tool that alerts investigators when a vehicle connected to an active investigation is spotted. Every alert is verified by an officer before any action is taken.
:format(webp))
PRODUCT SAFEGUARDS
PRODUCT SAFEGUARDS
Human Directed
A person verifies every alert. People make every decision
Built-In Transparency
Every search, alert, and access event is permanently recorded
Protected Data
All data is encrypted, protected, and never shared by default
:format(webp))
PRODUCT SAFEGUARDS
PRODUCT SAFEGUARDS
Human Directed
Built-In Transparency
Protected Data
Built Responsibly
This page is for community members who want to understand how Vehicle Intelligence works, what guardrails are in place, and why we made the choices we did. We believe how we build our products matters just as much as what we build. Our Responsible Innovation Framework is where that belief becomes practice.
What is Vehicle Intelligence
Vehicles are involved in a majority of serious incidents, yet identifying their role in an investigation remains one of the hardest challenges investigators face. A license plate is often only part of the story. Investigators also need context: what the vehicle looked like, where it was seen, and how it connects to other events.
Vehicle Intelligence is built on a simple belief: technology should help investigators uncover critical vehicle connections faster, while human judgment remains in charge. By combining license plate recognition, vehicle attribute detection, and real-time video across fixed and mobile cameras, Vehicle Intelligence helps investigators see, search, and respond to vehicle-related incidents from one connected platform.
The goal is safer communities and faster answers for the people affected by serious incidents. Vehicle Intelligence helps investigators develop leads more efficiently, spending less time searching across disconnected systems and more time on the decisions and community engagement that require human expertise. Investigators remain responsible for reviewing every detection, validating every match, and determining what action to take.
Vehicle Intelligence is shaped by community input and continuously evaluated to ensure it is used responsibly. The sections below outline the specific safeguards built into the product, and the community feedback that shaped each one.
What is Vehicle Intelligence
Vehicles are involved in a majority of serious incidents, yet identifying their role in an investigation remains one of the hardest challenges investigators face. A license plate is often only part of the story. Investigators also need context: what the vehicle looked like, where it was seen, and how it connects to other events.
Vehicle Intelligence is built on a simple belief: technology should help investigators uncover critical vehicle connections faster, while human judgment remains in charge. By combining license plate recognition, vehicle attribute detection, and real-time video across fixed and mobile cameras, Vehicle Intelligence helps investigators see, search, and respond to vehicle-related incidents from one connected platform.
The goal is safer communities and faster answers for the people affected by serious incidents. Vehicle Intelligence helps investigators develop leads more efficiently, spending less time searching across disconnected systems and more time on the decisions and community engagement that require human expertise. Investigators remain responsible for reviewing every detection, validating every match, and determining what action to take.
Vehicle Intelligence is shaped by community input and continuously evaluated to ensure it is used responsibly. The sections below outline the specific safeguards built into the product, and the community feedback that shaped each one.
:format(webp))
Key safeguards
The key themes from our community research on Vehicle Intelligence heavily inform how we design, develop, and roll out this technology. We evaluate these insights alongside rigorous ethical and inclusion standards to shape our product development process. The resulting safeguards are a meaningful reflection of that dialogue, purposefully designed to address the core issues that surface.
:format(webp))
Human Directed
Investigators must verify every alert before taking action. The system enforces strict guardrails, including mandatory search reasons, administrator-gated hotlists, and role-based access controls.
:format(webp))
Built-In Transparency
Every search, alert, and access event is recorded in a permanent audit log that tracks who accessed data, when, and why. The system preserves a strict chain of custody and offers an opt-in transparency portal for agencies to share policies and usage data directly with the community.
:format(webp))
Protected Data
Built with complete local ownership, agencies determine exactly how their encrypted and secured data is handled. The system prevents unauthorized third-party access and enforces a default 30-day retention unless data is tied to an active case.
Human Directed
WHAT WE HEARD FROM THE COMMUNITY:
Vehicle Intelligence systems should support investigations without replacing human judgment. Every search requires a documented reason, which builds in accountability from the start. There should be greater transparency around how vehicle data is collected, retained, and used, along with clear avenues for public oversight.
IMPLEMENTED SAFEGUARDS:
Human Verification of Every Alert: When a hotlist match is detected, an officer must validate the license plate number and state before any action is taken. Vehicle Intelligence surfaces information and it never acts on its own.
Search Reason Required: Every search requires an officer to enter a documented reason before results are returned. That reason is logged and auditable, creating a record of why the search was conducted.
Admin-Controlled Access: Agencies define who can search, who can view records, and who can create or modify hotlists. Those decisions stay with agency administrators, not Axon.
The System Identifies Vehicles, Not People: Vehicle Intelligence is designed to help locate vehicles connected to crimes. It does not identify drivers, predict behavior, or assign risk scores. Those boundaries are built into the product, not left to policy.
Human Directed
WHAT WE HEARD FROM THE COMMUNITY:
Vehicle Intelligence systems should support investigations without replacing human judgment. Every search requires a documented reason, which builds in accountability from the start. There should be greater transparency around how vehicle data is collected, retained, and used, along with clear avenues for public oversight.
IMPLEMENTED SAFEGUARDS:
Human Verification of Every Alert: When a hotlist match is detected, an officer must validate the license plate number and state before any action is taken. Vehicle Intelligence surfaces information and it never acts on its own.
Search Reason Required: Every search requires an officer to enter a documented reason before results are returned. That reason is logged and auditable, creating a record of why the search was conducted.
Admin-Controlled Access: Agencies define who can search, who can view records, and who can create or modify hotlists. Those decisions stay with agency administrators, not Axon.
The System Identifies Vehicles, Not People: Vehicle Intelligence is designed to help locate vehicles connected to crimes. It does not identify drivers, predict behavior, or assign risk scores. Those boundaries are built into the product, not left to policy.
:format(webp))
Built-In Transparency
WHAT WE HEARD FROM THE COMMUNITY:
Accountability should be built into the system from the start, not added after the fact. Every search and access event should be traceable and reviewable, with clear records of why data was accessed and by whom. A public-facing transparency portal should exist written in plain language, showing how the system is being used in their community, with pathways to report concerns or contest misuse.
IMPLEMENTED SAFEGUARDS:
Opt-In Transparency Portal: Agencies can choose to publicly disclose how they use Vehicle Intelligence, including data retention policies, program details, and deployment status, through a community-facing transparency portal.
Permanent Audit Log: Every search and investigative activity is recorded in a permanent log. This includes the reason an officer entered before conducting the search, creating a permanent record of not just what was searched, but why.
Chain-of-Custody Preservation: Audit records document who accessed data, when, and what investigative activity it was tied to. This ensures a complete and traceable record from the moment data is collected to how it is used.
Built-In Transparency
WHAT WE HEARD FROM THE COMMUNITY:
Accountability should be built into the system from the start, not added after the fact. Every search and access event should be traceable and reviewable, with clear records of why data was accessed and by whom. A public-facing transparency portal should exist written in plain language, showing how the system is being used in their community, with pathways to report concerns or contest misuse.
IMPLEMENTED SAFEGUARDS:
Opt-In Transparency Portal: Agencies can choose to publicly disclose how they use Vehicle Intelligence, including data retention policies, program details, and deployment status, through a community-facing transparency portal.
Permanent Audit Log: Every search and investigative activity is recorded in a permanent log. This includes the reason an officer entered before conducting the search, creating a permanent record of not just what was searched, but why.
Chain-of-Custody Preservation: Audit records document who accessed data, when, and what investigative activity it was tied to. This ensures a complete and traceable record from the moment data is collected to how it is used.
:format(webp))
Protected Data
WHAT WE HEARD FROM THE COMMUNITY:
Communities expressed concerns about who can access vehicle data, how long it is retained, and whether it could be shared beyond its intended purpose, including with immigration or civil enforcement entities. They called for strong agency control, clear retention limits, and protections against data being repurposed over time.
IMPLEMENTED SAFEGUARDS:
Local Ownership, Local Control: Agencies own their vehicle intelligence data outright. They set the rules for who can access it, how it is used, and when it is deleted.
No Sharing Without Authorization: Data is not shared by default. If an agency chooses to share access, it is explicit, invite-only, remains within the agency’s environment, and can be revoked at any time. Axon has no access to agency data. Agencies decide who sees their data, and when that access ends.
Automatic Deletion: Data is kept for 30 days by default, then automatically deleted unless an agency assigns it to an active case. Agencies can adjust this window to fit their local policies.
CJIS-Certified Security: All vehicle detection data is encrypted and stored in CJIS-certified cloud storage, the same security standard used for criminal justice records across the country.
Protected Data
WHAT WE HEARD FROM THE COMMUNITY:
Communities expressed concerns about who can access vehicle data, how long it is retained, and whether it could be shared beyond its intended purpose, including with immigration or civil enforcement entities. They called for strong agency control, clear retention limits, and protections against data being repurposed over time.
IMPLEMENTED SAFEGUARDS:
Local Ownership, Local Control: Agencies own their vehicle intelligence data outright. They set the rules for who can access it, how it is used, and when it is deleted.
No Sharing Without Authorization: Data is not shared by default. If an agency chooses to share access, it is explicit, invite-only, remains within the agency’s environment, and can be revoked at any time. Axon has no access to agency data. Agencies decide who sees their data, and when that access ends.
Automatic Deletion: Data is kept for 30 days by default, then automatically deleted unless an agency assigns it to an active case. Agencies can adjust this window to fit their local policies.
CJIS-Certified Security: All vehicle detection data is encrypted and stored in CJIS-certified cloud storage, the same security standard used for criminal justice records across the country.
:format(webp))
How Vehicle Intelligence works
The following scenario illustrates how Vehicle Intelligence works in practice. The specific capabilities available, and how agencies choose to use them, will vary by jurisdiction.
The scenario: A hit-and-run leaves a pedestrian seriously injured. Witnesses describe a dark blue sedan, but the license plate was only partially visible – something like "7K4" – and the driver fled before officers arrived.
STEP BY STEP
How Vehicle Intelligence works:
STEP BY STEP
How Vehicle Intelligence works:
STEP BY STEP
How Vehicle Intelligence works:
:format(webp))
When a hit-and-run happens, cameras already installed across the community can help piece together what happened. Fixed cameras on streetlights and poles read plates and capture vehicle attributes around the clock. That data is encrypted and stored in CJIS-certified cloud storage, before being automatically deleted after 30 days.
FAQs
Key terms
The following definitions explain important terminology referenced throughout this page.
Fixed Camera
Stationary infrastructure-mounted cameras (like Outpost and Lightpost) that provide constant surveillance of a specific area, as opposed to mobile or body-worn cameras
Mobile Camera
Unlike fixed infrastructure (like Outpost or Lightpost), mobile cameras are mounted within patrol vehicles to provide high-speed, AI-powered detection while officers are on the move
ALPR / ANPR
Automatic License Plate Recognition (called Automatic Number Plate Recognition in the UK). This technology automatically reads vehicle license plates and compares them against databases
Fleet 3
Axon’s AI-powered in-car video system for patrol vehicles, which includes integrated ALPR capabilities to capture data while officers are mobile
Outpost
A flexible, fixed camera that can be installed on various structures such as poles, buildings, or trailers. It is designed for versatile deployment and supports solar or AC power
Lightpost
A camera designed for rapid, large-scale deployment that mounts directly onto existing streetlights in under an hour, requiring no new trenching or specialized infrastructure
Fusus
The central aggregation platform that unifies feeds from fixed cameras, mobile Fleet 3 cameras, and third-party sources into a single operational map for real-time situational awareness
Axon Evidence
Axon Evidence is a secure, cloud-based system that helps law enforcement agencies store, manage, and review digital evidence, including videos captured on body-worn cameras
Vehicle attributes
Specific descriptive data captured by the cameras beyond just the license plate, including the vehicle's make, model, color, and other unique identifiers
Hotlist
A digital "watch list" of vehicles of interest (e.g., stolen cars or vehicles associated with AMBER alerts). When a camera detects a match, it triggers a real-time notification to officers
False alert
An instance where the system incorrectly identifies a vehicle as a "hit" on a hotlist
:format(webp))
:format(webp))