Skip to main content

Welcome to Axon's ACEIP US Law Enforcement Transparency Portal

Learn about our Axon Customer Experience Improvement Program (ACEIP) and upcoming ACEIP Basic use cases.

The Axon Customer Experience Improvement Program (ACEIP) enables U.S. Law Enforcement Agencies to actively contribute to the responsible evolution of Axon's public safety technologies. Through this voluntary program, participating agencies securely share transformed data to help improve product accuracy, reliability, and effectiveness while maintaining the highest standards of privacy and data protection. The ACEIP Transparency Portal serves as a public resource outlining each approved use case within the program, detailing what data is used, how it is protected, and the specific benefits it delivers to agencies and the communities they serve.

What is ACEIP?

ACEIP is a voluntary program that allows Axon to use Customer Content in a secure and controlled manner to accelerate Axon's development of technology, such as building and supporting automated features, aiming to increase safety within communities and efficiency in public and workplace safety. Access to Customer Content in Axon Cloud Services, voluntarily shared with us and securely protected, helps us create the best solutions possible.

Customers can participate in the ACEIP in two ways:

ACEIP Basic

A “limited” improvement program, which uses only transformed or de-identified data via Privacy-Preserving Techniques approved by Customers. No identifiers for the data are retained, e.g., Customer name, Customer ID, etc.

ACEIP Custom

Enables product improvements that cannot be accomplished with the limitations transformed data in Basic ACEIP.  Custom includes copies of certain segments or elements, or full copy of the Customer Content that has not been transformed via a privacy preservation technique. Custom is managed by specific agreements with the Customer. 

Please visit ACEIP to find more details on the program and participation opportunities.

Our Commitment to Transparency

This Transparency Portal provides information on how ACEIP Basic data is being used when collected from Law Enforcement Agencies in the US. It also provides detailed information on the purpose of such use, privacy preserving extraction technique (if applicable), and relevant data protection practices. As a reminder, participants may withdraw from ACEIP at any time. For more information on ACEIP or to submit a request to withdraw from ACEIP visit https://www.axon.com/aceip.

List of Use Cases

Draft One Customized Rules

Status: Active | Tier: ACEIP Basic | Region: US
September 2026

Customer Content used:
For eligible use of this feature, Axon may access the semantic rules learned from an agency's past draft one usage. The system only exports rules for review that are learned and generated from that process.

Detailed Description of Use Case

Supervisor Feedback Loop

Status: Active | Tier: ACEIP Basic | Region: US
September 2026

Customer Content used:
For eligible reviews, Axon's automated systems access the following within the customer environment. No Axon individual reviews transcript text, listens to audio, or views video as part of this use case:

  • The review's behavioral-signal annotations — the signal type, the supervisor's verdict (Confirmed / Rejected — binary, no "not sure") or suggested signal, the AI's 1–10 confidence, and the transcript turn position number (a number only, never the words).

  • The call type (from dispatch / CAD, e.g. "traffic stop").

  • The supervisor's reason for their rating

    Detailed Description of Use Case

Axon Assistant Translation Channel Query Redaction

Status: Active | Tier: ACEIP Basic | Region: US
August 2026

Customer Content used:
For eligible use of this feature, Axon may access the user’s query and limited context about the chat session. This may include the timestamp of the event, device type, device id and user’s native language.

Detailed Description of Use Case

ASR Corrections Training

Status: Active | Tier: ACEIP Basic | Region: US
August 2026

Customer Content used:
For eligible use of this feature, Axon’s automated systems may access the following data:

  • Records that show when a user saved a modified transcript, including limited identifiers needed to locate the related evidence.

  • Word-level corrections in transcript content and the associated contextual segment containing those corrections. No individual reads this transcript content.

  • Evidence audio for evidences that still have eligible short segments after correction screening and personal-information filtering. Short audio samples are used only to measure technical audio quality (for example, noise levels), and are not reviewed for their substantive content.


Detailed Description of Use Case

ECOM Natural Language Search Query Redaction

Status: Active | Tier: ACEIP Basic | Region: US
August 2026

Customer Content used:
For eligible use of this feature, Axon may access the user’s natural-language search query and limited context about the search event. This may include the partner (organization) identifier, a user identifier, a session identifier that groups turns of Natural Language Search interaction in the same session, a message identifier unique to the query, the verbatim query text as typed by the user (only sanitized version will be viewed by Axon personnel), and the timestamp the event was captured.

Detailed Description of Use Case

Form One Image Analysis and Image-Grounded Evaluation

Status: Active | Tier: ACEIP Basic | Region: US
August 2026

Customer Content used:
Form One is an AI-powered report assistance capability that helps officers complete structured fields in web-based forms using customer-provided inputs, including images. Images may include screenshots, scans, or photographs of driver’s licenses, identity documents, completed or partially completed forms, and other documents relevant to report completion.

Detailed Description of Use Case

TREND: Signal Sidearm Depersonalized Device Log Pipeline

Status: Active | Tier: ACEIP Basic | Region: US
August 2026

Customer Content used:
Axon will access anonymized SSR device logs containing activation broadcasts, holster/unholster events, operating-mode changes, firmware events, power events, error conditions, and periodic device-health heartbeats. The data includes the device category (firearm, OC spray, baton, or handcuffs), battery and calibration health, firmware version, and event timing.

Detailed Description of Use Case

Body-Worn Camera (BWC) Depersonalized Transcript Data for TREND

Status: Active | Tier: ACEIP Basic | Region: US
July 2026

Customer Content used:
Axon will access depersonalized BWC transcript content and metadata from recordings associated with TASER deployments, including: transcript text with personally identifying details removed or transformed; incident type and categorical metadata (e.g., level of force, weapons present, number of speakers); relative timing of spoken events; and linkage identifiers that connect a depersonalized transcript to a TASER deployment event without identifying individuals. 

Detailed Description of Use Case

Records (RMS) AI Search

Status: Active | Tier: ACEIP Basic | Region: US
July 2026

Customer Content used:
For eligible users of this feature, Axon may access the user’s search query and limited context about the search results. This includes a one-way hashed (pseudonymized) organization identifier, a pseudonymized user or session identifier, the sanitized query text, the sanitized AI interpretation of the query (the structured filters the system derived), whether the user clicked a result or abandoned the search, the rank of returned results, the type and broad category of result returned, a system confidence value, any thumbs-up/thumbs-down feedback the user gave on the AI interpretation or summary, and the date of the search. 

Detailed Description of Use Case

Case Agent User Query Redaction

Status: Active | Tier: ACEIP Basic | Region: US
July 2026


Customer Content used:
For eligible users of this feature, Axon may access the user’s query and limited context about the chat session. This may include the turn number in a chat session, the number of case evidences are present, whether or not the user gives thumbs feedback, and the date of the chat.
.

Detailed Description of Use Case

Case Agent Backtesting Use Case

Status: Active | Tier: ACEIP Basic | Region: US
July 2026


Customer Content used:
The quality-testing process accesses two types of data within the customer environment:

  • Evidence file: Axon's automated testing system retrieves sections of case. These files are not exported or saved; they are read within Case Agent's secure environment, used to generate test questions, and then discarded.

  • Synthetic test questions: Test questions are generated by an AI model based on the evidence excerpts.

Detailed Description of Use Case

Segmentation Mask Feedback Loop for RA

Status: Active | Tier: ACEIP Basic | Region: US
July 2026

Customer Content used:
Customer Content used is limited to anonymized, privacy-preserving information generated when a user edits AI redactions. Axon does not use raw video footage, facial features, raw pixels, or Personally Identifiable Information. Instead, Axon receives anonymized structural outlines of the scene, redacted scene-layout descriptions scrubbed of identifying details, redaction box coordinates and IDs, the action taken on the AI box, and the general object category when available, such as “monitor screen” or “notebook,” for the purpose of improving AI model performance.

Detailed Description of Use Case

Response-to-Resistance (R2R) & Use-of-Force (UoF) Form Data for TREND

Status: Active | Tier: ACEIP Basic | Region: US
June 2026


Customer Content used:

  • Axon will access data from the R2R/UoF form itself including anonymized event details, depersonalized subject details, depersonalized officer details, use-of-force circumstances, tactical use of force facts, and outcomes, linked TASER device telemetry already attached to the report such as which probes were deployed.


Detailed Description of Use Case

Redaction Assistant Model Training

Status: Active | Tier: ACEIP Basic | Region: US
April 2026


Customer Content used:

  • Axon will access a subset of customer videos together with the associated redaction mask information needed to train and evaluate the object detection models. This includes the set of masks that users created, updated, or deleted, and the associated frame-level information needed to interpret those masks


Detailed Description of Use Case

Evidence-Case Matching Quality Evaluation

Status: Active | Tier: ACEIP Basic | Region: US
April 2026


Customer Content used:

  • Axon will access the transcriptions generated from your uploaded video and audio evidence, including body-worn camera footage, in-car video, interview room recordings, and other audio/video. For each transcription, the associated date, time, and location will also be accessed. Axon will also access your case information, including case descriptions, case notes, case tags, case status, and the record of which evidence items are associated with which cases. No video, audio, or image files are accessed - only the text transcriptions and case information described above.


Detailed Description of Use Case

Attribute Search User Query Redaction Feedback Loop

Status: Active | Tier: ACEIP Basic | Region: US

March 2026

Customer Content Used:

  • Axon may access the user’s search query and limited context about the search results. This includes a pseudonymized organization identifier, a pseudonymized user identifier, the sanitized query text, whether the user clicked a result or abandoned the search, the confidence scores associated with returned results, and the date of the search.

  • Detailed Description of Use Case

Form One User Suggestion Rejection Feedback Loop

Status: Active | Tier: ACEIP Basic | Region: US

March 2026

Customer Content Used:

  • Axon may access information about each submitted form and the outcome for each evaluated field. The exported data model consists of a pseudonymized Organization ID, pseudonymized User ID, pseudonymized Form ID, pseudononymized Field ID, Entity Type, Error Category, and Date. No raw form content, transcripts, dictation, or field values are extracted, and only categorical outcomes are exported.

  • Detailed Description of Use Case

Brief One AI Search User Query Redaction Feedback Loop

Status: Scheduled | Tier: ACEIP Basic | Region: US

March 2026

Customer Content Used:

  • Axon may access the user’s search query and limited context about the search results. This may include whether the user clicked a result or abandoned the search, the rank of returned results, the type of result returned, broad category labels associated with the result, a confidence value, and the date of the search.

  • Detailed Description of Use Case

Differential Privacy Transcript Generator

Status: Active | Tier: ACEIP Basic | Region: US

March 2026

Customer Content Used:

  • Axon accesses automatically generated transcripts derived from body-worn camera audio recordings. These transcripts may represent full interactions captured during body-worn camera recordings. This use case involves access only to automatically generated BWC transcript text.

  • This use case does not involve access to raw audio recordings, video footage or images

  • Detailed Description of Use Case

Measuring success & topics in Axon Assistant (General & Agency Info)

Status: Active | Tier: ACEIP Basic | Region: US
March 2026

Customer Content Used:

  • When a user submits a question to Axon Assistant, the system briefly processes limited request metadata to generate a response. This includes the question text, the channel it was asked in (e.g., General Info or Agency Info), the time of the request, a request ID, a partner ID used only to separate agency data, and a session ID used to group queries from the same session.

  • Detailed Description of Use Case

Training Discriminator Models to Detect Synthetic Transcripts
Transcript Quality Tool

Status: Active | Tier: ACEIP Basic | Region: US
February 2026


Customer Content Used:

  • Axon uses existing transcript text generated from body-worn camera audio to help build and refine this AI tool. During this process, the tool learns how to distinguish between higher-quality and lower-quality transcripts based on overall patterns, not individual cases or content.

  • Detailed Description of Use Case

AI Auto-Tagging Shadow Testing Feedback Loop

Status: Active | Tier: ACEIP Basic | Region: US
February 2026


Customer Content Used:

  • Axon will access the following information from a completed Auto-Tagging workflow: the body-worn camera transcript associated with the incident (used temporarily and not stored), the list of incident categories available for the agency, the category or categories suggested by Evidence.com, the final category or categories selected by the user, event timing information (including incident time and category selection time), system information such as the model and prompt version used, and an agency identifier, which is immediately converted into a non-reversible internal identifier; Axon will use this data solely to evaluate and improve the quality of AI-generated incident category suggestions.

  • Detailed Description of Use Case

Audio Alerts Feedback Insights

Status: Active | Tier: ACEIP Basic | Region: US
February 2026


Customer Content Used:

  • Short transcript snippets associated with Audio Alerts, along with customer feedback provided through thumbs up or thumbs down buttons. This feedback is used only to improve the alert system and is not an evaluation of officer performance, use of force, or decision-making. For example, a thumbs up indicates the alert was helpful or appropriate and a thumbs down indicates the alert was not helpful or did not reflect the situation accurately.

  • Detailed Description of Use Case

Detect Error Cases of ALPR and SOR

Status: Active | Tier: ACEIP Basic | Region: US
February 2026


Customer Content Used:

  • Axon will access ALPR data entries from customer database, which contains: (1) images captured by Fleet / Outpost camera, (2) ALPR read result and metrics, (3) user-correction, and (4) state level location of the reads.

  • Detailed Description of Use Case

Reference Database and End-to-End Quality Improvement

Status: Active | Tier: ACEIP Basic | Region: US
December 2025

Customer Content Used:

Redaction Assistant Model Training Input Signal Analysis
User masks & foundational model insights

Status: Active | Tier: ACEIP Basic | Region: US

December 2025

Customer Content Used:

Transcript Metadata Extraction

Status: Active | Tier: ACEIP Basic | Region: US
December 2025

Customer Content Used:

  • Text transcripts generated from BWC recordings for customers participating in Axon’s Customer Experience Improvement Program.

Detailed Description of Use Case

Edit Analysis for Draft One

Status: Active | Tier: ACEIP Basic | Region: US
August 2025


Customer Content used:

  • Draft transcripts from your Axon Evidence tenant

  • Summary of edits and corrections made to the draft transcript


Detailed Description of Use Case

Auto-Transcribe Platform Improvement

Status: Active | Tier: ACEIP Basic | Region: US
July 2025


Customer Content used:

  • Transcripts Customers corrected and their associated evidence audio


Detailed Description of Use Case

Feedback loop for Agencies to evaluate and improve their TASER programs

Status: Active | Tier: ACEIP Basic | Region: US
July 2024


Customer Content used:

  • TREND form and, when available, attached TASER logs


Detailed Description of Use Case

Automatic License Plate Recognition (ALPR) System Improvement

Status: Active | Tier: ACEIP Basic | Region: US
October 2022


Customer Content used:

  • License plate images captured by a users Fleet Camera and the associated ALPR Read Record such as date and time of the capture, and geographical location


Detailed Description of Use Case

Auto-Transcribe Platform Improvement Phase 2.0

Status: Active | Tier: ACEIP Basic | Region: US
March 2023


Customer Content used:

  • Audio in your Axon Evidence tenant and associated transcripts

  • Metadata associated with the evidence files such as titles, descriptions, and notes created by you and associated with any

    images or other types of media or multimedia in your Axon Evidence tenant

  • Customer entity and User Service Interaction Data: This is data about your interactions with our devices or Axon Evidence, such as

    device/client information, click/UI events, session length etc.


Detailed Description of Use Case

Comments or Further Questions?

We value your feedback and are committed to addressing any questions or concerns about ACEIP for law enforcement agencies, please contact [email protected]. Looking for another ACEIP region or sector? Please visit our ACEIP page at https://www.axon.com/aceip.

Resources