Flock Gunshot Detection: CEO and Website Tell Two Stories

Flock CEO Garrett Langley has a simple description of what his company's microphones listen for: gunshots, and nothing else. Flock's own website describes something broader. That gap came to the surface at TechCrunch Disrupt this week, where Langley defended the product onstage while protesters gathered outside the San Francisco venue.

The exchange matters beyond one company. It shows how hard it is to pin down what an always-on sensing system actually does when the people selling it describe it in different ways depending on the audience.

What Langley said onstage

Flock is a surveillance company best known for license plate reading cameras. Its audio product, called Gunshot Detection, uses microphones placed across neighborhoods to pick up gunfire and other ambient sounds. The stated goal is to give public safety agencies "earlier awareness" of incidents.

Speaking with TechCrunch editor-in-chief Connie Loizos, Langley explained the reasoning behind the product. "Most people don't call 911 when they hear a gunshot, and so that's why we built the product," he said.

He then described how the scope briefly grew. Some customers had the microphones installed in parking lots and asked whether they could also be alerted if someone was being abducted or harmed. Langley said the idea seemed logical, so Flock built a prototype and tested it with a couple of customers. Concerns followed. Langley admitted the company had made mistakes, but said it had listened.

"There's only a couple customers that were prototyping it with us, and we said, 'OK, great, it's just going to do gunshots, plain and simple,'" he said. If a gun goes off, the system notifies. Otherwise, in his words, "that's all it's going to listen to."

When pressed on the fact that Flock's marketing claims more than that, his answer was short: "It's just gunshots."

What the website says

Flock's product pages tell a different story. They say the audio systems can pinpoint gunfire instantly, but also "detect street takeovers early to stop chaos before it escalates" and "identify severe crashes in real time so help arrives faster." Another line promises to "expand awareness beyond gunfire with alerts for car crashes, fireworks, and sideshows." Sideshows and street takeovers are terms used in the US for informal car gatherings that occupy public roads.

The FAQ goes further. It states that "in addition to gunshots, Flock Gunshot and Audio Detection can detect fireworks, sideshows, and other disruptive community events." It also says the company's "detection library is continually expanding to include emerging threats that are relevant" to what individual customers need.

After the event, Flock spokesperson Paris Lewbel offered a clarification. The technology, Lewbel said, "is primarily used to detect gunfire but is also capable of detecting fireworks, sideshows, and potential car crashes." According to the spokesperson, Langley meant to stress that Flock's Audio Detection "is not designed to monitor conversations or analyze speech and does not currently detect human screaming or distress."

The "screaming" episode

That last point refers to an earlier controversy. When Flock announced Gunshot Detection last year, it said the system would listen for "screaming." Privacy groups objected to a product that listened for human voices. Flock changed the wording on its website to "sounds of human distress," and later said it had removed the distress detection feature entirely from the customers who had tried it.

Langley framed much of the criticism as misinformation, arguing that critics often misunderstand how Flock's products work. On scale, he said Flock's microphone count is comparable to ShotSpotter, the best-known gunshot detection vendor, which had about 25,000 microphones as of 2024, according to Wired.

A widening backlash

The interview comes as public sentiment turns against Flock and other license plate reader companies. Privacy advocates, including the ACLU, have raised concerns about misuse, and some members of the public have gone as far as destroying the company's cameras. Texas and Florida have moved to end their relationships with Flock, and Amazon's Ring has ended its partnership with the company. Similar unease about fast-moving technology shows up in polling, such as a survey finding that most Americans want AI to slow down.

Our Take

The core issue here is not whether Flock's microphones can hear fireworks. It is that the CEO, the spokesperson and the website describe the product's scope in three different ways. For a system that sits in public spaces and listens continuously, the definition of what it detects is the main safeguard the public has. If that definition shifts depending on who is talking, trust erodes quickly.

The phrase "continually expanding" detection library deserves the most attention. It suggests the capabilities of deployed hardware can change through software updates, without new microphones or obvious public notice. The screaming episode showed how fast a feature can be prototyped with customers before outside scrutiny catches up.

This fits a wider pattern of AI-assisted tools in policing raising questions about accuracy and oversight, as seen when an AI model sent a false murder tip to police. It is worth watching whether Flock publishes a fixed, public list of detected sound categories, and whether more states or partners follow Texas, Florida and Ring in stepping away.