Teens Sue School Police After Doritos Bag Triggers AI Gun Alert
Four Black Baltimore County high school students were handcuffed and searched at gunpoint after an AI system mistook a Doritos bag for a gun. District officials canceled the alert within two minutes, but the untrained principal dispatched police anyway. Six patrol cars responded with weapons drawn. The teens have filed a federal lawsuit alleging unlawful search and detention, citing that over 99% of the district's AI alerts were false positives.


Short answer: Four Black Baltimore County high school students were handcuffed and searched at gunpoint after an AI system mistook a Doritos bag for a gun. District officials canceled the alert within two minutes, but the untrained principal dispatched police anyway. Six patrol cars responded with weapons drawn. The teens have filed a federal lawsuit alleging unlawful search and detention, citing that over 99% of the district's AI alerts were false positives.
teens flagged by AI gun detection Doritos
Four Black high school students were sharing a bag of Cool Ranch Doritos outside their Baltimore County school after football practice in October 2025 when the campus AI gun detection system flagged one of them as a potential shooter. Within two minutes, district officials reviewing the linked camera feed recognized the snack bag and canceled the alert. That should have been the end of it. Instead, the principal, who had never received training on the Omnilert system, texted a school resource officer fifteen minutes later to dispatch police, noting in the same message that the object might not be a gun. At least six patrol cars converged on the scene with lights flashing. Some officers arrived with weapons drawn. The teens froze, cooperated, and were ordered to their knees on the concrete, handcuffed, and searched while officers separated them and told the others they were guilty by association. Only after finding no weapon did the students learn an algorithm had mistaken a chip bag for a firearm. One officer reportedly muttered that the AI cameras were stupid. Another thanked them for not running, adding that would have been a real problem. As they walked away, an officer joked they should be careful with their chip bags.
students sue school district over AI gun alert
The students, Taki Allen, Yui-Osa Ogbeide, Devon Thompson Jr., and Keith Winfield III, have filed a federal lawsuit against the school district, the police department, Omnilert, and individual officials. Their complaint alleges unlawful search and detention in violation of constitutional rights and seeks an injunction halting the district's use of Omnilert until protocols exist to prevent police responses to canceled alerts. The Lawyers' Committee for Civil Rights Under Law represents them. The suit also cites a staggering figure: over 99 percent of the district's AI gun alerts were false positives, a claim that underscores how often the system cries wolf. Damon Hewitt, the committee's president, described the combination of blind trust in unproven technology and broken communication between agencies as a toxic mix that puts children in life-or-death situations.
investigation clears Omnilert finds human failures
An investigation by the Office of the Inspector General for Education cleared Omnilert of bias but exposed a cascade of human failures. The delegation of authority process at Kenwood High School did not stop the principal from acting on a canceled alert. The OIG recommended biannual training for all staff and a reduction in the number of people who receive gun alerts so that internal controls are followed and police only respond to verified threats. Without those changes, the report warned, the loss of notification controls could endanger students, staff, and responding officers alike.
The fallout for the teens has been severe. Allen no longer feels safe waiting for rides after school and now pays for private car services. Winfield, who lives near the campus, calls his mother repeatedly after practice to confirm she is coming. All four report lasting anxiety, fear, humiliation, and physical injuries from the handcuffs. Rumors that a gun was actually found have damaged their reputations. Their mothers have demanded systemic change so no other child faces armed police over a snack.
AI public safety deployment warning false positives
For anyone building or deploying AI in public safety, this case is a warning. A vendor's marketing claim of near-zero false positives means nothing if the human layer, training, escalation protocols, verification steps, is missing or broken. The district adopted Omnilert early and touted it as the largest deployment of its kind, yet the principal operating it had zero training. The system worked as designed: it flagged an anomaly and humans canceled the alert. The catastrophe happened because a human ignored the cancellation and triggered an armed response anyway. Builders must design for that reality. They must assume alerts will be wrong, that staff will be untrained, and that communication will fail. Guardrails need to be technical, not just procedural, automatic lockouts when alerts are canceled, mandatory verification before police dispatch, audit trails that cannot be bypassed.
Readers responsible for AI procurement or oversight should demand evidence of false-positive rates in their specific environment, not vendor brochures. They should require contractual obligations for training and protocol certification before go-live. They should insist on independent red-teaming of the full human-AI loop, not just the model. And they should treat any system that can summon armed force to a schoolyard as safety-critical infrastructure, subject to the same rigor as medical devices or aviation software. The Doritos bag was harmless. The response was not. The difference was not the algorithm. It was the absence of a system designed to keep humans in control.
Frequently asked questions
What triggered the police response to the four Baltimore County students in October 2025?
An Omnilert AI gun detection system flagged one student holding a Cool Ranch Doritos bag as a potential shooter. District officials canceled the alert within two minutes after recognizing the snack bag on camera, but the untrained principal texted a school resource officer fifteen minutes later to dispatch police anyway, noting the object might not be a gun.
Who are the students suing the school district and police department?
The plaintiffs are Taki Allen, Yui-Osa Ogbeide, Devon Thompson Jr., and Keith Winfield III, four Black high school students at Kenwood High School. They were handcuffed, searched, and forced to kneel on concrete by at least six patrol cars with weapons drawn after the canceled AI alert.
What does the federal lawsuit allege and seek?
The lawsuit alleges unlawful search and detention violating constitutional rights. It names the school district, police department, Omnilert, and individual officials. The students seek an injunction halting the district's use of Omnilert until protocols exist to prevent police responses to canceled alerts. The Lawyers' Committee for Civil Rights Under Law represents them.
What did the Office of the Inspector General investigation reveal?
The OIG cleared Omnilert of bias but exposed human failures: the principal had zero training on the system, the delegation of authority process failed to stop action on a canceled alert, and over 99 percent of the district's AI gun alerts were false positives. The OIG recommended biannual training for all staff and reducing alert recipients to restore internal controls.
What lasting effects have the students reported since the incident?
All four students report lasting anxiety, fear, humiliation, and physical injuries from handcuffs. Allen now pays for private car services instead of waiting for rides. Winfield repeatedly calls his mother to confirm pickup. Rumors that a gun was found have damaged their reputations, and their mothers demand systemic change.
Source: Ars Technica
Related posts

Anthropic IPO Raises $81.3 Billion Amid Governance Questions
Anthropic completed its IPO on November 19, 2026, raising $81.3 billion-the largest ever, surpassing SpaceX. The AI lab, founded by ex-OpenAI staff prioritizing safety, adopted a Public Benefit Corporation structure with a Long-Term Benefit Trust holding controlling voting rights alongside founders. However, governance experts question the trust’s independence and AI expertise, noting its four trustees lack deep AI backgrounds and a recent trustee joined Anthropic as an executive, rais

Anthropic Goes Public on Nasdaq at $111.50 Per Share
Anthropic began trading on Nasdaq under ticker ANT on November 19, opening at $111.50 per share, above its $83 IPO price. The debut valued the AI company at roughly $1.5 trillion, the largest public offering on record. Despite reporting an $80.7 billion net loss for the first half of 2026, revenue reached $16.3 billion. Founders retain control through a dual-class structure. CEO Dario Amodei pledged to prioritize AI safety over profit pressure.

Google Rolls Out Computer Use for Mac Gemini
Google has rolled out a computer use feature for its Gemini macOS app, letting the AI assistant operate across local files, desktop programs, and the web to complete multi-step tasks. Available now in English for U.S. users, the system requires per-app consent and displays a live preview of every action. The launch precedes Apple's upcoming macOS 27.3 restrictions on full disk access for AI agents.