Police rely on the community to call 9-1-1 if gunshots are fired, but only 20% of incidents are ever reported on average. This creates a situation where police departments have a large data gap that makes it difficult to be able to effectively “serve and protect” when it comes to gun violence. SoundThinking’s ShotSpotter acoustic gunshot detection technology fills the gap with a network of acoustic sensors that can detect, locate, and alert police to nearly all gunshot incidents. The system is in operation in more than 180 cities and is used by police to: 1) be able to respond to a higher percentage of gunfire incidents; 2) improve response times to crime scenes to better aid victims and find witnesses; and 3) help police locate key evidence to identify and investigate suspects.

ShotSpotter uses an array of acoustic sensors that are connected wirelessly to ShotSpotter’s centralized, cloud-based application to reliably detect and accurately locate gunshots using triangulation. Each acoustic sensor captures the precise time and audio associated with impulsive sounds that may represent gunfire. This data is used to locate the incident and is then filtered by sophisticated machine algorithms to classify the event as a potential gunshot. Acoustic experts, who are located and staffed in ShotSpotter’s 24×7 Incident Review Center, ensure and confirm that the events are indeed gunfire. They can append the alert with other critical intelligence such as whether a fully automatic weapon was fired or whether there are multiple shooters. This entire process takes less than 60 seconds from the time of the shooting to the digital alert popping onto a screen of a computer in the 911 Call Center or on a patrol officer’s smartphone or mobile laptop.

ShotSpotter is used in more than 180 cities and is highly regarded by law enforcement as a critical component of gun violence prevention and reduction strategies. ShotSpotter protects a wide range of city types and sizes ranging from urban metropolitan cities such as Las Vegas and New York City; to medium-sized cities such as Boston, Denver, and Oakland; and small cities with populations less than 50,000 such as Richmond, CA and Pleasantville, NJ.

Gunshot detection by itself is not a panacea for gun violence, but if used as part of a comprehensive gun crime response strategy, it can contribute to a reduction in response times, help to save the lives of GSWs, and improve evidence collection rates. Please visit ShotSpotter Facts for more details.

The ShotSpotter system is highly accurate at detecting outdoor gunshots. From 2019-2021 the system had a 97% aggregate accuracy rate across all of our customers, including a very small customer-reported false positive rate of less than 0.5%.

No. Strict controls and policies ensure that sensors only trigger on loud explosive or impulsive sounds that are likely gunfire – not human voices, even loud ones. An independent privacy audit conducted by NYU’s Policing Project concluded that the risk of voice surveillance is extremely limited.

ShotSpotter does not and will not listen for the sounds of human voices. No community member’s information is obtained or used during any part of the detection or alerting process. Our technology focuses solely on gunshots, not voices. Learn more about community privacy protections here.

ShotSpotter provides police with a tool to respond to and investigate gunshot incidents in a more precise area compared to the 9-1-1 system to avoid canvassing entire blocks or neighborhoods. There is zero data supporting the claim that ShotSpotter puts police on high alert or creates dangerous situations. Rather, ShotSpotter equips police officers with more information than they might typically have when arriving to the scene of a gunshot incident, and they arrive at the scene more situationally aware. Additionally, ShotSpotter provides intelligence that allows police to coordinate a safe and efficient response, requiring fewer resources, and in a way that can enhance community trust.

Right now, we do not have a solution for individual community members. However, you can write a letter to your local elected officials to request ShotSpotter for your neighborhood.

Regardless of whether your city has deployed ShotSpotter, you should still call 9-11- and notify the police. You may not be in a ShotSpotter coverage area as it typically doesn’t cover an entire city. Over 80% of gun crimes go unreported and your call could help save lives. If you’re in a city that has not deployed ShotSpotter, consider reaching out to your local police department or local city council to express your concerns about public safety and to see if they have any plans to implement gunshot detection technology.

SoundThinking offers:

• CrimeTracer™, the #1 law enforcement search engine that enables investigators to search through more than 1B structured and unstructured data across jurisdictions to obtain immediate tactical leads, leverage advanced link analysis to make intelligent connections, link NIBIN leads to reports, suspects, and other entities, and more.
• CaseBuilder®, a smarter investigation management platform that enables investigative teams to efficiently organize and store case information, track and prioritize assignments, optimize resources, submit materials to prosecution, and more.
• CaseBuilder® Crime Gun, the first-of-its-kind gun crime tracking and analysis case management tool that enables agencies to better capture, track, prioritize, analyze, and collaborate on incidents that involve firearms.
• ResourceRouter™ , is a resource management tool that automates the planning of directed patrols and provides transparency around engagement activities across an entire jurisdiction, daily.

There are multiple technology and policy protections in place to protect against audio surveillance:

• Sensors are placed high above the street typically on building or streetlights to avoid street level sounds and the microphones used are not specialized in any way (i.e. everyday cell phone quality)
• The system is tuned to listen for loud impulsive sounds that are gunshots or similar to gunshots (fireworks, car backfires) and takes no action on other sounds that would include street level sounds or human voices
• The sensors store a limited amount of audio locally and that audio is automatically purged every 30 hours
• Sensors are triggered and an incident created only when 3 or more sensors hear the same loud impulsive sound and can verify a location. This creates an incident and sends a short audio snippet to the ShotSpotter Incident Review center. The snippet has the gunfire and 1 second of audio prior to and after the gunfire to establish an ambient noise level. Audio snippets are typically only a few seconds long unless there is a gun battle.
• The company’s culture of privacy protection builds in proper measures when new features or processes are contemplated

• The Policing Project of NYU Law School conducted an independent audit of ShotSpotter’s gunshot detection system to evaluate the privacy risk. In July 2019 they published a report that stated that there is an extremely low risk of human voice surveillance.
• The Oakland Privacy Advisory Commission reviewed ShotSpotter in November 2019 under one the most stringent local surveillance ordinances in the country. The Commission unanimously approved the continued use of ShotSpotter
• In 2020, the ShotSpotter system passed review by the San Francisco Privacy Advisory and Surveillance Board.

• There is no ability to listen to live streamed audio from the sensor
• If the system misses a gunfire incident, police may contact the company to see if there is any audio or location evidence. In this case, only authorized ShotSpotter personnel with proper credentials can access sensor audio to search. Their search is limited to the 30- hour sensor storage timeframe. The agency must provide evidence of a shooting in order for ShotSpotter personnel to initiate access to a sensor such as a victim, witness or shell casings. Searching is done visually first, not by listening, to identify when impulsive sound events occurred. Once these events are noted, a short portion of the audio is downloaded for auditory review. An audit trail tracks who accessed the sensor and who requested the audio search.

• While an individual sensor could potentially hear a human voice, that sound would be purged within 30 hours with no ShotSpotter employee, police department customer or other 3rd party having the ability to hear it unless an incident was created
• Human voices are not loud enough to trigger sensors and are rarely heard as part of an incident under review. The person would have to be screaming very loudly above the loud sound that initiated the incident and the audio last only a few seconds.
• In the past extended audio was available to police agencies and prosecutors for court cases. This is no longer possible under new policy and technology controls.

• No personally identifiable information is associated with an incident.

• Audio at the sensor level is purged every 30 hours
• If an incident is created and sent to our Incident Review Center the short audio snippet is stored permanently for evidentiary purposes as well as to train the machine learning model.
• Per the NYU’s Policing Project recommendation, we only store one second of pre- and post-incident audio.

The technology and privacy policies have evolved over time. Up until approximately 2012, police departments had direct access to extended audio and some agencies used that to get auditory evidence of sounds prior to and after a gunshot. Significant new protections have been added so that police no longer have access to extended audio and only are able to hear the short audio snippets that are produced when the system and human reviews believe a sound is a gunshot.

Data is protected through encryption both at rest (typically AES 256) and in transit (typically via TLS 1.2 with MTLS where appropriate).

These documents are available upon request to validated customers with a current MNDA. Please contact our customer support or sales team to initiate the process.

SoundThinking maintains compliance with standards such as SOC2, HIPAA, CJIS, CIS Baselines, and NIST 800-53.

Includes hardened configurations, asset tracking, monitoring, maintenance, and patching.

No. Sensors do not record or trigger based on speech. Only a 3-second clip may be captured if associated with a gunshot.

Uses SHA-256 hashing, a zero-trust model, and only returns anonymized, aggregated data.

AI is governed by internal policies, bias evaluation, audits, human oversight, and transparency.

Through certified staff, ongoing education, AI-enhanced threat detection, and evolving best practices.

Email: security@soundthinking.com with detailed info and supporting evidence, if available.