A Chicago-based startup, SplitSec, is conducting trials of a new gunshot detection application designed to leverage the microphones already present in Android smartphones. This technology aims to create a network of users that can alert others to shootings, potentially offering a more widespread and accessible solution compared to fixed sensor installations.
The app, developed by SplitSec, a company founded in Chicago in 2025, uses an AI model that runs entirely on the user's phone. According to SplitSec founder and CEO Sri Deivasigamani, the goal is to make people aware of gunfire incidents sooner. "As I observed events around us, one of the things that was striking is any time there is a gunfire incident, people are unaware, and they stand there for several minutes not knowing what happened," Deivasigamani told Fox News Digital. "And so the idea was how can we make people aware sooner?"
Two no-cost trials are currently underway in Chicago's 21st and 42nd wards, with discussions about expanding to additional wards. Deivasigamani noted that early feedback from trial users has highlighted issues such as false positives on phones with highly sensitive microphones and battery drainage, which SplitSec engineers are addressing. He also reported that on July 27 and 28, the app successfully detected gunshots that users confirmed hearing.
SplitSec's approach differs from the gunshot detection technology previously used by Chicago, which relied on fixed sensors. Deivasigamani explained that advances in smartphone AI and processing power have made this mobile-based detection possible. The app is designed to activate only upon detecting gunshot sounds, similar to how voice assistants respond to specific phrases.
Once the app detects a potential gunshot, it prompts the user to confirm. If confirmed, SplitSec can distribute an alert to other users within a one-mile radius. The company stated that if two or more users confirm the same incident, their location can be triangulated, helping individuals move away from danger.
Privacy concerns have been raised regarding the use of personal devices for surveillance. Deivasigamani clarified that the audio analysis is performed on the smartphone itself. Audio clips are only sent to the company if a user reports a false positive, and this data is used to train the AI model and improve accuracy. "We really believe such an app needs to earn people's trust and needs to be built in such a way that it protects privacy," he said.
Chicago previously contracted with SoundThinking, formerly known as ShotSpotter, but ended that agreement in 2024, discontinuing police alerts from the city's sensor network. SplitSec is not currently collaborating with the Chicago Police Department, and its alerts are intended for app users. However, Deivasigamani indicated that if a partnership were to form with city agencies, the gunfire alert data could be shared with law enforcement.