AI · 20 November 2026
AI gun-detection false alarm over Doritos bag leads to teen lawsuit
Teenagers held at gunpoint after an AI gun-detection system flagged a Doritos bag have sued their school system and local police, according to Ars Technica.
What happened
Teenagers who were held at gunpoint after an AI gun-detection system flagged a crisp packet have filed a lawsuit against their school system and local police. Ars Technica reports that the students say the encounter left them traumatised.
The alert was triggered by a Doritos bag, which the detection system appears to have mistaken for a weapon. According to the lawsuit as reported, the system's alerts were overwhelmingly wrong: more than 99% of the school system's AI gun alerts were false positives.
The case puts the operation of automated threat detection in schools under legal scrutiny, and it centres on the people who were stopped and the consequences of a misidentification.
Why it matters
For organisations deploying AI in safety-critical, public-facing settings, this case shows how a system's accuracy can look sound on paper while its real-world error rate creates serious harm. A detection tool that is wrong almost every time it fires does not just produce noise. It changes how staff respond, how students experience the building, and how much trust the community places in the institution.
It also highlights that the experience of an AI decision is shaped by the human process around it: who verifies an alert, how quickly, and what happens before officers act on it. Those design choices sit with the operator, not the algorithm.
By the numbers
- Over 99% of the school system's AI gun alerts were reported as false positives.
- November 2026 is the publication date of the Ars Technica report, based on its URL.
The Renascence take
Most coverage will focus on the lawsuit and the Doritos bag. The more important finding is the 99% figure, which suggests the problem was not a rare edge case but a system whose alerts were rarely reliable. Any safeguard that relies on a machine's output as the trigger for a human response has effectively handed the experience of those affected to the model's error rate.
The lesson is not that AI has no place in school security. It is that any alert with a 99% false-positive rate should never be allowed to start a response that puts a child at gunpoint. Operators should map the full chain from detection to human verification to physical action, and insert deliberate friction before the most harmful step. They should also measure false positives as a customer-experience metric, because the people who bear them are the ones who live with the outcome.
Sources
This briefing was written by our Newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.
FAQ
Questions we get on this topic
More in AI
Stay ahead of CX
Get the signal, not the noise.
The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.


