In a curious incident that unfolded at Kenwood High School in Essex, Maryland, an AI security system mistakenly identified a bag of Doritos as a firearm, leading to swift police action and a temporary lockdown. This event raised significant discussions on the capabilities and limitations of AI in critical safety roles. The occurrence, which took place last week, involved the Omnilert system, a sophisticated AI-powered tool designed for weapons detection. The system scanned live footage from cameras positioned around the school and erroneously flagged the recognizable orange snack bag as a potential firearm, prompting an immediate alert at 7:04 PM. Within minutes, trained safety personnel reviewed the video and quickly nullified the initial alert after confirming it was a false positive. However, the alarm had already triggered a police dispatch, which resulted in officers responding to the scene to assess the situation. Baltimore County Public Schools (BCPS) confirmed the incident, expressing the intent to enhance the protocols surrounding the use of the AI system. In a joint statement with local officials, BCPS recognized the efficiency of their trained staff but acknowledged that a communication breakdown had occurred. Baltimore County Councilman Julian E. Jones Jr. emphasized the necessity for ongoing vigilance, leading to decisions for annual retraining of staff on the security protocols surrounding the technology. The incident highlights not only the challenges of integrating advanced technology in public safety but also the essential need for human oversight and adaptability when implementing such systems. Following the event, BCPS is implementing various changes, including requiring that school administrators directly notify specific safety directors for any security concerns. Additionally, police are set to reinforce procedures for school resource officers after school hours to ensure better communication in the future. As AI technologies continue to evolve, the Doritos false alarm serves as a poignant reminder of the risks associated with automation in security contexts. Despite the errors of AI systems, officials have reassured the public of their commitment to the technology, citing its importance as part of a multi-faceted approach to school security. The Omnilert system, designed to analyze the characteristics of objects in real-time video feeds, remains in use across all public schools in Baltimore County, despite the recent blunder. Officials noted that there have been successful identifications of genuine threats in the past, underscoring that while errors may occur, the AI system can play a significant role in ensuring safety in educational environments. Experts in the field of AI and public safety suggest that while these systems are usually reliable, they are not infallible. The phenomenon of false positives is well-documented, with AI systems facing particular difficulties distinguishing between benign objects and potential threats. As a result, organizations that utilize such technology must maintain rigorous training programs and ensure that clear communication protocols are established between staff and law enforcement. Looking ahead, the Baltimore County incident may serve as a case study for other educational institutions and security agencies seeking to adopt similar technology. The mistakes made in Essex stand as a cautionary tale of the importance of balancing technological reliance with human intervention. This balance is critical to optimize both safety and efficiency in school environments.