Cool Ranch, Hot Mess: Students Sue After AI Gun Detector's Epic Failure
HOME›NEWS›LAW
LAW

Maryland Teens Sue Police After AI Mistook Doritos Bag for Gun

Four Maryland teens are suing Baltimore County police, the school board, and several staff after an AI gun detector at Kenwood High School flagged them. Officers held one student at gunpoint and handcuffed him before finding the 'weapon' was a bag of Cool Ranch Doritos. The 31-page complaint says officers knew the system was unreliable.

Bearing Arms|November 18, 2026|12h ago|5 min read|ORIGINAL SOURCE ↗

Teens Sue Baltimore County After AI Scanner Mistook Doritos for a Gun

Four Maryland teenagers are suing the Baltimore County Police Department, the county school board, and several officers and school employees after an AI gun detection system falsely flagged them for carrying a firearm on campus. The alert came from Omnilert Gun Detect at Kenwood High School last October. Officers responded with guns drawn, forced students to their knees on concrete, handcuffed them, and searched them. One student was held at gunpoint while police looked for a weapon that did not exist. The object the software identified as a gun was a bag of Cool Ranch Doritos. The parents filed a 31-page complaint that targets the police, the district, and the vendor that sold the software.

Key Details

The facts in the complaint are specific, and they matter because they show the sequence of decisions, not just a technology glitch.

  • Plaintiffs: Jerrie Davis and Lamont Davis, parents of the four students, filed the complaint.
  • Defendants: Baltimore County Police Department, the Baltimore County school board, individual officers, and school district employees. The complaint also targets the company behind the system.
  • Technology: Omnilert Gun Detect, an AI system that flags possible firearms from camera footage.
  • Location and timing: Kenwood High School, October of the previous year.
  • Police conduct: Officers exited their vehicles with weapons drawn and did not explain the stop. Officers searched the students and found no weapons before anyone mentioned the Omnilert alert.
  • Officer remarks: According to the complaint, one officer told the students that AI was not reliable. Another made disparaging remarks about the cameras. The parents argue officers relied on the alert anyway.
  • Principal conduct: Kenwood principal Catherine Smith allegedly texted staff about the incident with four laughing-crying emojis.
  • Apologies: The lawsuit says no school district employee has apologized to the students.
  • Vendor position: The company says a human reviewer should have checked the alert before police responded, and that the software otherwise worked as designed.
  • District position: Baltimore County Public Schools Superintendent Myriam Rogers told local media the program depends on human verification and performed as intended by raising an alert for people to examine.

Background

School gun detection software took off after the 2018 Parkland shooting in Florida. Districts across the country bought camera-based systems that scan feeds for weapons and push alerts to staff or police. Vendors sell these products on the promise of faster response. The weak point is always the same: an alert is only as good as the review behind it. Software does not sort a bag of chips from a pistol in poor lighting or at an odd angle. A person has to make that call before anyone draws a gun. In this case, the complaint says the review step either did not happen or was skipped in favor of an immediate armed response.

The legal framework for this kind of stop predates the software by decades. In Terry v. Ohio (1968), the Supreme Court allowed brief stops based on reasonable suspicion. In Florida v. J.L. (2000), the Court held that an anonymous tip that a young man carried a gun was not enough to justify a stop. The tip did not show the person was doing anything illegal, and that ruling still governs how courts view unverified gun reports. New York State Rifle & Pistol Association v. Bruen (2022) changed the Second Amendment analysis for carry laws. It did not change the Fourth Amendment rules for police stops, but it made clear that gun-related government action must match historical tradition.

Why It Matters for Gun Owners

This case reaches every lawful carrier, not just students. If an algorithm can trigger a felony-style response against a teenager carrying a snack, the same logic can point at an adult with a holstered pistol in a parking lot or a sporting goods store. Gun owners in Maryland and every state that uses school or public camera detection should understand that an alert is not a finding of fact. The first questions that matter are simple. Who reviewed the image? How long did that review take? Did officers confirm a weapon before drawing?

Practically, carriers should expect that a camera alert can produce a sudden, armed contact. Keep your hands visible, avoid sudden movements, and comply calmly with commands. Do not reach for a bag, pocket, or waistband when officers approach. If you are stopped, say you are carrying only if asked, and keep your mouth shut beyond that. Document the encounter as soon as you can. Video from a phone or a bodycam request can make the difference in court.

Local officials should also face pressure. Ask your school board how their detection system is configured, who reviews alerts, what the response protocol requires, and whether officers must verify a weapon before using force. Those answers are public record in many districts, and they tell you whether a contractor's software is running a school or a school is running a contractor's software.

DownRange Analysis

The defense the district offers is weak. Saying the system worked because a human was supposed to check it puts the failure on whoever skipped that step, and the complaint says officers responded as if the alert were already proven. Under Florida v. J.L., an unverified report of a gun does not automatically justify a seizure. An AI alert sits even further from a witness's first-hand observation. A court applying Fourth Amendment reasoning should ask what the officers actually knew, and the complaint alleges they knew the software was unreliable. Bruen does not directly govern police stops, but its demand for careful historical and legal reasoning around firearms should make courts skeptical of gun-based responses built on unverified data.

For the market, expect school districts to demand vendors carry liability and accuracy guarantees, and expect vendors to push verification costs onto schools. Gun owners should watch whether Maryland lawmakers respond with standards for detection systems, and whether anti-gun politicians who loudly oppose civilian carry say anything about armed government responses triggered by software. Right now, do three things. Know your state's stop-and-carry rules. Practice calm, visible compliance drills. Support public records requests that reveal how any gun detection system in your area is actually used. Read the original reporting at Bearing Arms for the full complaint details.

ORIGINAL SOURCE
This editorial was written by DownRange based on the original article. Read the primary source for additional detail.
READ ORIGINAL ↗
TAGS
ai-gun-detectionmarylandbaltimore-countyschool-safetyfourth-amendmentpolice-misconductcivil-rights-lawsuit
SHARE:X / TWITTERFACEBOOK