DHS · DHS-2572 · record R1878

Acoustic Signature AI for Gunshot Detection

Department of Homeland Security · CBP

Overview

Development stage: PilotSource: b) Pilot – The use case has been deployed in a limited test or pilot capacity.
High-impact designation: Presumed high-impact, determined notSource: b) Presumed high-impact, but determined not high impact
Impact justification: This system detects sounds that are associated with sounds of gunfire. Only events classified as High Confidence are then sent to users as a detection. Users then review the audio of the event to determine the accuracy of the detection and to review the location of the detection on an associated map. If the audio review is deemed relevant by agent and/or system users, agents may be sent to investigate further and the message is sent to field personnel of potential gunshot activity. It may be warning of activity south of the border where notifications made enhance safety of officers/agents in the region. Any further actions taken will be based on what the agents encounter when they arrive on scene. The AI associated with this system is not a principle basis for a decision or action.
Start date: 2025-09-08 (day precision)Source: 2025-09-08T00:00:00
Withheld from public reporting: Not reported

Mission

Topic area: Law Enforcement
Operational functions: No function tag matched Derived by keyword rules; see methodology

Problem the AI is intended to solve

The AI will be used to make confidence determinations between non gunshot acoustic activity and actual gunshot activity to reduce false alerts to agents monitoring the User Interface.

Expected benefits

Gunshot and UAS detection notifications can be used by Agents for enhanced situational awareness in their area of operations. By utilizing this AI learning technology, they will have high confidence of alerts of a detection, with the specific type, as well as a pinpointed GPS location within 3-meters.

System outputs

Agents will get a real time notification via text with a link to the location on maps as well as a link to an audio clip recording of the event. With Agent confirmation that the event was correctly identified, the AI utilizes all of the information for future cases. This real time notification grants Agents the ability to know what is happening right now and anticipate what might happen next in their environment. Increasing agent and officer safety in the geographic region.

Technology

AI classification: Classical / predictive MLSource: Classical/Predictive Machine Learning: Models trained on data to make predictions or classifications based on identified patterns or relatio…
System name(s): Not reported
Custom-developed code: No
Public source code: Not reported

Sourcing

How it was built: Purchased from vendorSource: a) Purchased from a vendor
Vendor (as reported): Invariant Corporation
Vendors (standardized): None on the standard list

A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.

Data and privacy

Involves PII: No
Privacy Impact Assessment: Not reported
Authorization to Operate: No
Demographic features: Not reported
Training and evaluation data: Initial training was conducted by the vendor using vendor obtained audio. Additional data was captured during deployments with other customers for further refinement.
Federal Data Catalog: Not reported

Governance

0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.

Pre-deployment testingNo answer reported

AI impact assessmentNo answer reported

Independent reviewNo answer reported

Ongoing monitoringNo answer reported

Operator trainingNo answer reported

Fail-safeNo answer reported

Appeal processNo answer reported

User and public consultationNo answer reported

Potential impacts and how they were identified

Not reported