Texas charter school deploying AI software that can detect guns via camera feed; how it works

Yahoo

21-05-2025

AUSTIN (KXAN) — Artificial intelligence is integrating into the daily lives of Texans more and more as the state is believed to be poised as one of the nation's leaders in AI and technology.
One company, founded in 2018 by a group of Navy SEALs, is using AI to expand safety against the threat of firearms in schools.
ZeroEyes uses AI to detect the presence of guns spotted by surveillance cameras. It was the first AI-based gun detection video analytics platform to earn the full U.S. Department of Homeland Security SAFETY Act Designation.
RELATED | Schools are turning to artificial intelligence to spot guns
On Tuesday, Ki Charter became the first charter school system in Texas to deploy the software.
'We serve a student population that requires an exceptional level of care and security, and I was initially skeptical of ZeroEyes, but after seeing the technology in action, I knew we needed it,' Superintendent Jerry Lager said in a press release. 'Our largest campus spans 65 open acres, making security a complex challenge. ZeroEyes provides that extra layer of protection, ensuring our students and staff are in a safe environment.'
Ki Charter has campuses in San Marcos, Liberty Hill, Austin, Temple, Arlington, Fort Worth, Denton, and Kingwood, and serves over 6,500 students annually. According to its website, the school system 'is the premier educational provider for students who reside in residential facilities (RFs) and day centers.'
Ki Charter 'reduces learning gaps to help students learn behavioral management strategies and improve their communication skills,' per its website.
Co-founder Sam Alaimo said it took about two years to build the algorithm that ZeroEyes uses. The software is layered onto existing digital security cameras to detect guns. If a gun is identified by the software, images are shared with the ZeroEyes Operations Center, which Alaimo said was modeled after the military's Tactical Operations Center.
The operations center is the industry's only U.S.-based, fully in-house operations center, which is staffed 24/7 by specially trained U.S. military and law enforcement veterans, according to the company.
If experts in the operations center determine a threat and notify law enforcement and school security, they include information like a visual description, gun type, and last known location when sending those alerts.
Alaimo broke down the process:
'[The] algorithm says, 'I think it's a gun.' Human verifies it's a gun, hits dispatch. Client gets the alert.'
That happens in just a few seconds.
'So in the real world, on schools, subway platforms, shopping malls, you name it, once that gun is exposed in front of a client camera, within about three to five seconds, that client will get that image, as well as local law enforcement. And what we're trying to do is cut through the fog of war,' Alaimo said. 'It's like a pure situational awareness tool.'
Alaimo said ZeroEyes helped prevent a potential tragedy in Texas last year when three people showed up at an elementary school with AK-47s.
'This is what we're in the business of detecting,' Alaimo said. 'We detect guns every single week at this point, and it's an interesting position to be in, because when we send an alert like this, and people get arrested, there is no mass shooting. So we can't quantify how many mass shootings didn't happen as a result of our software working.'
Alaimo said the company was founded after the 2018 Parkland, Florida, mass shooting at Marjory Stoneman Douglas High School.
'We were all kind of commiserating on, hey, we need to do something meaningful again. We want to be of service again,' Alaimo said. 'And in 2018, the Parkland shooting happened, and that was when America did a bit more introspection after one of these massive shootings, and decided that we wanted to do something about it.'
'So we all quit our jobs and decided to found ZeroEyes with the intent to keep kids safe from mass shootings,' Alaimo continued. 'And we started the company, bootstrapped it. It took about two years to build the algorithm, and we finally had it ready to go to market in 2020.'
Alaimo said the software had 'a lot of good traction with K-12' at the time, but the COVID-19 pandemic derailed the company's plans because of schools going into lockdown. At that point, ZeroEyes went into the commercial space, which Alaimo said was 'a good move.'
'We found that, you know, identifying guns in subway platforms and in grocery stores and on Navy bases actually made us better at identifying guns in a school environment,' he said. 'So… there's a lot of ways that training the algorithm in very diverse environments strengthens it.'
In 2021, ZeroEyes founders were able to get back to their passion of 'keeping kids safe.'
Now, the software is deployed across 46 states in K-12 schools, in higher education, and in various commercial spaces.
The cost of implementing the software depends on the needs of the client. Pricing is primarily based on the number of camera streams that will have ZeroEyes installed, according to its website.
'It usually goes between 20 and $60 per camera stream per month, higher if it's a small [number of cameras], what, like, 25 camera streams for one year, and much, much lower if it's a large number of cameras over, over three or five years,' Alaimo explained.
Alaimo said because it's a new technology that not many schools have the money budgeted for, ZeroEyes has grant writers on staff to help schools find the money to pay for it. The company also works with legislators and will lobby governments, he said.
'Money shouldn't be the reason why a school does not have our software. So for that reason, we actually have grant writers on staff to help schools find the money,' Alaimo said. 'And that's one of the reasons we try to work with legislatures as often as we can, because once they know we exist, then it's like, all right, how do we get money to get them out there?'
Copyright 2025 Nexstar Media, Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.