Beyond Metal Detectors: How Schools Are Rethinking Security
in the Age of AI and Behavioral Threat Assessment

The Threat Landscape: Progress and Persistence 

The statistical picture of school violence in 2025 offers genuine grounds for cautious optimism, and equally genuine grounds for continued urgency. Omnilert’s 2025 Gun Violence Statistics Report, the most comprehensive annual analysis of school shooting data compiled from the K-12 School Shooting Database and multiple federal sources, documented that school shootings decreased 30 percent from 2024 to 2025, deaths and injuries in K-12 school shootings fell 26 percent, and mass shootings declined 19 percent nationally (Campus Safety Magazine, 2026; Omnilert, 2026). The 2024-25 school year recorded 254 shooting incidents on K-12 campuses — a significant drop from the approximately 330-incident average of the preceding three years (Building Security Services, 2026). These are real improvements, and they reflect the cumulative effect of multi-year investments in prevention, detection, and response infrastructure across thousands of school districts.

What the aggregate decline obscures is the persistence of catastrophic individual incidents and the structural geography of the threat. Two-thirds of all school shootings over the past six decades occurred outdoors — predominantly in parking lots and in front of school buildings — meaning that the security infrastructure most schools have concentrated inside their perimeters addresses approximately one-third of the actual threat environment (Omnilert, 2026). The December 2025 Brown University shooting, which killed three and wounded nine, reinforced that higher education campuses face a categorically different security challenge from K-12 environments — attackers who move between multiple buildings across open campuses in ways that no fixed security checkpoint can address (Omnilert, 2025). Student exposure to school shootings has nearly tripled since 1999, rising from 19 to 51 per 100,000 students despite declining incident counts in 2025, reflecting the concentration of violence in larger events (Building Security Services, 2026). Moreover, swatting, the deliberate filing of false emergency threats to trigger large-scale law enforcement responses, increased significantly in 2025, costing taxpayers over $1 billion and straining exactly the emergency response infrastructure that genuine threats require (Omnilert, 2026). The threat has not been solved. It has shifted, and the security architecture responding to it must shift accordingly.

The Technology Layer: What AI Can and Cannot Do

The school security technology market in 2026 is delivering capabilities that were operationally impossible five years ago. Next-generation weapons detection systems — millimeter-wave and multi-sensor platforms that have largely displaced traditional walk-through metal detectors at well-resourced facilities — can screen hundreds of students simultaneously, distinguish between metallic objects with dramatically greater accuracy, and minimize the bottlenecks that made traditional screening impractical at large schools during peak arrival periods (Campus Safety Magazine, 2026a). Video analytics platforms can now differentiate between a holstered weapon and a brandished weapon in a parking lot, identify the early behavioral indicators of a fight in a cafeteria before physical contact occurs, and trigger automated lockdown protocols within seconds of confirmed weapon detection (Campus Safety Magazine, 2026a). Sound analytics systems have been detecting aggression through vocal pattern analysis for several years and are increasingly integrated with video platforms to provide compound alerting — two independent sensor types confirming the same behavioral anomaly before any alert is generated (Campus Safety Magazine, 2026a).

These capabilities are genuinely valuable as force multipliers for trained security professionals. The operative phrase is force multipliers. Campus Safety Magazine’s 2025 year-in-review analysis of AI’s impact on campus security technology concluded with a finding that the Maryland and Tennessee incidents had already demonstrated in practice: “it is critical to pair these tools with strong policies, comprehensive training, and responsible oversight to prevent unintended harm” (Campus Safety Magazine, 2025, para. 9). The AI gun detection platform that identifies a potential firearm in a parking lot has completed approximately 10 percent of the security task. The remaining 90 percent — verifying the detection, assessing the specific threat it represents, determining the appropriate response, communicating with law enforcement, managing the movement of hundreds of students, and making the split-second judgments that determine whether a situation is contained or escalates — requires trained human professionals whose decision-making cannot be automated. School security expert Guy Bliesner, whose 2026 analysis is cited by both Campus Safety Magazine and NBOA, identified this as the defining tension in contemporary school security: the need to balance people with technology in a market environment where vendor incentives consistently push institutions toward the technology side of that balance (NBOA, 2026).

The technology vendor ecosystem has a structural interest in the perception that its products constitute solutions. Bliesner’s 2026 analysis is pointed on this: “vendors and manufacturers will continue to market themselves as ‘solutions providers,'” and this framing “has and will continue to foster the perception for too many that the purchase of one of these systems IS the solution” (Campus Safety Magazine, 2026b, para. 4). It is NOT. A millimeter-wave screening system with no trained operator to assess ambiguous results produces exactly the kind of false positive that led to the detention of a Maryland student after an empty bag of chips. An AI-powered behavioral monitoring platform with no human reviewer to contextualize flagged content creates exactly the conditions that led to a Tennessee 13-year-old spending a night in jail. The technology is not the problem. The assumption that technology replaces human judgment is the problem — and it is an assumption that costs institutions, as well as sometimes students, at exactly the moment the system is supposed to protect them.

Behavioral Threat Assessment: The Human Foundation That Technology Cannot Replace

The most consequential development in school security over the past decade is not a sensor platform or an analytics algorithm. It is the near-universal adoption of Behavioral Threat Assessment and Management, the systematic, human-led process of identifying, investigating, and managing concerning behavior before it escalates to violence. A January 2025 nationally representative survey of 1,746 K-12 school principals, conducted jointly by the RAND Homeland Security Operational Analysis Center and the U.S. Secret Service National Threat Assessment Center, found that 97 percent of public K-12 schools in the United States had a BTAM team or equivalent program as of the 2024-25 school year — up from 42 percent in 2016 and 64 percent in 2020 (Campus Safety Magazine, 2026c). This adoption trajectory is the clearest available evidence that the field has moved, in both theory and practice, from reactive hardware to proactive human intelligence as the primary school safety architecture.

BTAM’s operational logic is grounded in a finding that the U.S. Secret Service’s National Threat Assessment Center has documented consistently across decades of research into targeted school violence: attacks are rarely impulsive. They are planned. They leak. Prior to virtually every documented school attack, the perpetrator exhibited concerning behaviors — communications, social media activity, direct statements to peers, changes in conduct, expressions of grievance — that were observable to trained professionals and, more often than not, to students who did not know how or whether to report what they had seen (Campus Safety Magazine, 2024). BTAM is the institutional infrastructure for collecting those observations, assessing their meaning through a structured multi-disciplinary process, and intervening before the behavior trajectory reaches a point of no return. As the SchoolSafety.gov definition captures it, the primary goal of BTAM is “to evaluate the difference between making a threat and posing a threat”, a distinction that requires clinical judgment, contextual knowledge of the individual, and cross-functional collaboration that no algorithm can approximate (Campus Safety Magazine, 2024, para. 3).

The RAND/NTAC survey’s findings about implementation quality are as important as its adoption statistics, and considerably more sobering. Despite 97 percent adoption, only 53 percent of schools had written policies explicitly establishing the scope of their BTAM program. Only 49 percent had written standard operating procedures for BTAM teams. Only 23 percent had dedicated program funding. Only 20 percent involved parents directly in intervention planning, despite research establishing parental involvement as a significant factor in successful case outcomes (Campus Safety Magazine, 2026c). These gaps are not technology gaps. They are human capital gaps — insufficient training, insufficient staffing, insufficient institutional investment in the professionals who make the system work. A BTAM team whose members have not been trained in structured threat assessment methodology, who lack the time to meet weekly as a full team to review cases, and who have no dedicated funding to sustain their operations is not a functional BTAM team. It is a compliance checkbox that provides the appearance of prevention infrastructure without the substance.

The Wood County Schools initiative in Ohio provides a concrete model of what genuine BTAM investment looks like in operational practice. Led by the county’s Alcohol, Drug Addiction and Mental Health Services Board, the program unites educators, counselors, public health agencies, and law enforcement in a single BTAM framework, using a centralized case management system that allows every relevant professional to observe early indicators of student distress, share information securely, and intervene with coordinated care before challenges escalate (Campus Safety Magazine, 2025). The program’s explicit philosophy — that “a student in distress is not a disciplinary concern but a young person in need of care” — reflects the clinical orientation that distinguishes effective BTAM from punitive surveillance (Campus Safety Magazine, 2025, para. 8). The technology supporting this program represented by the case management platform, the secure communication tools, the data integration with conduct and mental health records, is essential infrastructure. The program’s effectiveness depends entirely on the educators, counselors, law enforcement officers, and mental health professionals who exercise judgment within it.

The Human Capital Imperative

The Building Security Services 2026 school security statistics analysis identifies a gap that no technology investment can close: only 61 percent of schools have dedicated security staff on-site at least once a week — a decline from 65 percent in 2019-20 — despite 93 percent of schools using security cameras and 97 percent controlling building access (Building Security Services, 2026). The divergence between technology adoption and human staffing is not a resource allocation anomaly. It reflects a systematic undervaluation of human security professionals relative to hardware and software — an error whose consequences are visible in every incident where technology flagged a threat that no trained human was positioned to assess and act on appropriately.

School Resource Officers represent the most direct expression of the human security imperative in the K-12 environment. Sworn law enforcement officers with full arrest authority, formal criminal justice training, and a long-term campus assignment that allows them to build the kind of relational intelligence — who is struggling, who is isolating, whose behavior has shifted in the past two weeks — that no surveillance camera can develop, SROs function as both a security presence and a BTAM asset (Omnilert, 2026). The Omnilert analysis of SRO effectiveness is careful: the research evidence on SROs and juvenile crime is mixed, and the best outcomes are consistently associated with programs in which the SRO’s role is clearly defined, youth-focused training is ongoing, and the SRO is embedded in the school’s broader prevention architecture rather than operating as a standalone enforcement presence (Omnilert, 2026). The question, as the analysis frames it, is not “SRO or no SRO” but whether the district has “a student-centered, clearly defined and measurable safety program that protects students and supports a positive school climate” in which the SRO is one integrated component (Omnilert, 2026, para. 9).

Wisconsin’s Speak Up, Speak Out anonymous tip line — which received 6,946 tips in the 2024-25 school year, a 32 percent increase from the prior year — illustrates how human reporting infrastructure, properly resourced and embedded in a BTAM framework, generates actionable intelligence that no passive surveillance system produces (Campus Safety Magazine, 2025). The tip line works because it is operated by professionals who can assess credibility, investigate reports, and connect the dots between multiple low-level indicators that, individually, appear innocuous but collectively describe a student on a dangerous trajectory. The technology — the anonymous reporting platform, the case management system behind it — is the delivery mechanism. The value is generated by the trained professionals who act on what it delivers. Sandy Hook Promise’s “Say Something” program, which has trained 31 million individuals in bystander reporting and intercepted 18 credible threats in 2025 alone, operates on the same human intelligence principle: empower people who have observational access to concerning behavior, and provide trained professionals to receive, assess, and act on what they report (Building Security Services, 2026). 

The Integrated Architecture: Technology in Its Proper Role 

The emerging school security consensus in 2026, reflected across Campus Safety Magazine’s trend analyses, the ASIS International ANSI-approved School Security Standard, and the PASS Guidelines version 7, establishes an integrated architecture in which technology and human capital each occupy their appropriate role (Campus Safety Magazine, 2026b). Technology handles the tasks that technology does better than humans: continuous monitoring of large physical areas, consistent application of detection algorithms without fatigue or distraction, rapid processing of large sensor data volumes, and automated alert generation that compresses the time between detection and human notification. Human professionals handle everything that technology cannot: contextual judgment about whether a detected anomaly represents a genuine threat, knowledge of the specific individual whose behavior has been flagged, clinical assessment of threat trajectory, coordination of multi-agency response, and the relational work — with students, families, mental health providers, and law enforcement — that constitutes genuine prevention. 

Campus Safety Magazine’s 2026 technology trend analysis makes the sequencing explicit: advances in AI detection and video analytics “must be assessed holistically to ensure [they] strengthen the safety” architecture rather than distort it and the assessment standard is whether each technology investment “strengthens the safety of the entire school community” within a mission-focused framework, not whether it addresses the vendor’s marketing proposition (Campus Safety Magazine, 2026a, para. 5). The NBOA’s summary of Bliesner’s 2026 analysis frames the human-technology balance as the defining challenge of the current moment: schools are making genuine progress on BTAM integration and prevention architecture, while simultaneously facing vendor and legislative pressure to mandate specific technologies without regard to whether those technologies fit the specific facility’s threat profile, budget, or human capital capacity (NBOA, 2026). One-size solutions, in school security as in any complex operating environment, produce security theater — the appearance of protection without its substance. 

Conclusion 

The schools that are demonstrating the most effective security outcomes in 2026 share a common architecture: a robust BTAM program, well-funded and staffed with trained professionals who meet regularly, assess cases systematically, and intervene early; human security personnel — SROs, security officers, counsellors — who are embedded in the school community with sufficient relational access to observe what cameras cannot; anonymous reporting infrastructure that empowers students and staff to surface concerning behaviour before it reaches a threshold of visible danger; and technology — AI detection, video analytics, access control, emergency communications — deployed in the specific roles where it genuinely outperforms human capability, integrated with the human professionals who interpret and act on its outputs. What these schools do not have is the belief that any of those technology systems, however sophisticated, can substitute for the trained, experienced, contextually informed human professionals who make every other element of the architecture functional. The Maryland bag-of-chips incident and the Tennessee 13-year-old are not edge cases to be explained away. They are the predictable outcome of deploying sensors without judgment, and the clearest available argument for why the human capital investment in school security is not a line item to be optimized away in favor of a system that markets itself as a solution.

References 

Building Security Services. (2026). School security statistics 2026. https://www.buildingsecurity.com/statistics/school-security/ 

Campus Safety Magazine. (2024, December 17). Part 1: School and university behavioral threat assessment and management (BTAM) basics. https://www.campussafetymagazine.com/insights/part-1-school-and-university-behavioral-threat-assessment-and-management-btam-basics/164991/ 

Campus Safety Magazine. (2025, November 26). Wood County Schools pilot behavioral intervention program to enhance safety. https://www.campussafetymagazine.com/insights/wood-county-schools-pilot-behavioral-intervention-program-to-enhance-safety/174721 

Campus Safety Magazine. (2025, December 5). Wisconsin’s Speak Up, Speak Out program sees surge in school safety tips. https://www.campussafetymagazine.com/cs/wisconsins-speak-up-speak-out-program-sees-surge-in-school-safety-tips/175210 

Campus Safety Magazine. (2025, December 22). 2025 year-in-review: AI’s impact on campus security technology. https://www.campussafetymagazine.com/insights/2025-year-in-review-ais-impact-on-campus-security-technology/175543/ 

Campus Safety Magazine. (2026a, January 7). 5 school security technology trends to watch in 2026. https://www.campussafetymagazine.com/insights/5-school-security-technology-trends-to-watch-in-2026/175807 

Campus Safety Magazine. (2026b, January 12). School security trends: Predictions and aspirations for 2026. https://www.campussafetymagazine.com/insights/school-security-trends-predictions-and-aspirations-for-2026/175978/ 

Campus Safety Magazine. (2026c, January 7). 97% of schools have behavioral threat assessment and management teams. https://www.campussafetymagazine.com/insights/97-of-schools-have-behavioral-threat-assessment-and-management-team/176191/ 

Campus Safety Magazine. (2026, January 20). Gun violence statistics: School shootings decreased 30% last year. https://www.campussafetymagazine.com/insights/gun-violence-statistics-school-shootings-decreased-30-last-year/176739/ 

Campus Safety Magazine. (2026, January 14). 4 campus security trends to watch in 2026. https://www.campussafetymagazine.com/insights/4-campus-security-trends-to-watch-in-2026/176281 

National Business Officers Association. (2026, January 12). School safety trends in 2026. https://www.nboa.org/net-assets/news/view/school-safety-trends-in-2026 

Omnilert. (2025). School shootings 2025: Key data, incidents, and trends. https://www.omnilert.com/blog/school-shootings-2025 

Omnilert. (2026). School shootings 2026: Data, definitions and key events. https://www.omnilert.com/blog/school-shootings-2026 

Omnilert. (2026). 2025 gun violence statistics in the United States. https://www.omnilert.com/gun-violence-statistics 

Omnilert. (2026). What is a school resource officer and their role? https://www.omnilert.com/blog/school-resource-officer 

RAND Corporation & U.S. Secret Service National Threat Assessment Center. (2025). The state of behavioral threat assessment and management in K-12 public schools. https://www.secretservice.gov/sites/default/files/reports/2025-11/The-State-of-Behavioral-Threat-Assessment-and-Management-in-K-12-Public-Schools.pdf