AI Surveillance in Schools Expands Despite Reliability Gaps

American schools are deploying artificial intelligence surveillance systems at an accelerating pace, often without evidence that the technologies work as advertised or consideration of their impact on student trust and equity.

The consequences can be severe. In 2025, a Baltimore high school student carrying a bag of chips triggered an AI weapon-detection alert. Before a school resource officer could review the false alarm, police had already been called. Officers pointed guns at 16-year-old Taki Allen, forced him to his knees, and handcuffed him. "The first thing I was wondering was, was I about to die?" Allen later said.

Why it matters

The uneven deployment of school surveillance technology threatens to deepen existing inequities in American education. Research shows these systems are concentrated in schools serving low-income students and students of color—the same populations already subject to disproportionate disciplinary action and police contact. Without transparency requirements, independent validation, or meaningful community input, districts risk trading student trust for unproven security theater.

Surveillance Tools Proliferate Without Proof

The range of AI surveillance in schools now extends far beyond hallway cameras. Some districts have installed bathroom audio monitors that manufacturers claim only flag sounds like gunshots or distress calls without storing recordings. Yet one student successfully hacked his school's device to eavesdrop in real time and create fake alerts, exposing serious security vulnerabilities.

One Florida district went further, sharing student grade and attendance data with police to generate a list of "potential future criminals." The algorithm flagged students who had committed no offense, assigning risk scores based on academic performance. After 30 organizations mobilized opposition, the school board ended the data-sharing agreement.

Deployment Follows Familiar Patterns

A 2025 study found surveillance technologies are more prevalent in large, urban high schools with higher proportions of racial and ethnic minority students and low-income families. This unequal exposure matters because predictive policing tools can create self-fulfilling prophecies—students labeled as "high risk" receive more police attention, which can drive the very disengagement and distrust the systems claim to prevent.

Many of these tools emerged after school shootings, with companies like ZeroEyes launching AI gun-detection software specifically in response to the 2018 Marjory Stoneman Douglas tragedy. Yet students report that increased surveillance makes them less likely to trust adults at school or discuss mental health struggles—the very conversations research shows are most effective at preventing violence.

What Districts Should Require

Experts recommend several baseline protections before schools adopt surveillance technology:

Independent validation. Vendors should provide published error rates and third-party testing results before implementation, not after districts discover false-positive rates on live students.

Data limits. Schools should prohibit using academic and disciplinary records to generate predictive risk scores and strictly limit sharing student data with law enforcement.

Community engagement. Districts should seek input from students, parents, and teachers before deployment, not after incidents occur. Baltimore County Public Schools surveyed the community about continuing its AI system only after the Allen incident.

Transparency requirements. Schools should disclose when they collect biometric data or enroll students in facial recognition systems, even if families cannot opt out.

Only 10% of parents attended a school board meeting in the past year, according to survey data, though barriers like work schedules make participation difficult for many families. Written testimony offers an alternative for those unable to attend in person.

The fundamental tension remains: surveillance is not the same as safety. A school where every student feels watched is not necessarily a school where every student feels safe.

These findings were detailed in a Brookings Institution analysis by Rashawn Ray and Rebekah Boitey examining the expansion of AI surveillance in American schools.