Evolv AI Weapons Detection System Allegedly Misrepresents Accuracy, Leading to School Security Gaps

Evolv's AI-powered weapons scanners were advertised as superior to metal detectors but reportedly failed to detect actual weapons while generating excessive false positives. The FTC alleged that misleading claims about the system's accuracy and speed contributed to schools relying on unreliable detection, with at least one incident in October 2022 involving a missed knife that resulted in a student being stabbed.

Risk classification

• Primary risk domain: 7 AI system safety, failures, & limitations
• Primary risk subdomain: 7.3 Lack of capability or robustness

The primary issue is the AI system's lack of capability and robustness, as it failed to reliably detect weapons while generating high rates of false alarms on benign objects.

Additional risk subdomains

• 5.1 Overreliance and unsafe use: School districts placed inappropriate trust in the unproven AI system for critical safety, leading to a false sense of security and a failure to prevent a physical attack.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The incident was caused by the Evolv Express AI system failing to perform its core function of detecting weapons, which was deployed post-training in schools.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: The AI system is used as a safety component in critical public infrastructure (schools and public transit) and for law enforcement/security screening, which has significant implications for public safety and fundamental rights.

AI system and alleged parties

• AI system: Evolv Express (Evolv Technology)
• AI purpose: Threat Detection; Identification
• Behaviour type: Autonomous
• Alleged developer: Evolv Technology
• Alleged deployer: Charlotte Mecklenburg School District
• Alleged harmed parties: teachers at Charlotte Mecklenburg Schools, students at Charlotte Mecklenburg Schools, security officers at Charlotte Mecklenburg Schools, School administrators at Charlotte Mecklenburg Schools, Parents of students at Charlotte Mecklenburg Schools

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Substantial. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Minor
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Substantial, indirect Substantial
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Minor
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Substantial
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Physical

Reported: Yes, the report explicitly describes physical harm to a student caused indirectly by the system's failure.

Indirectly caused: In October 2022, a student was stabbed with a seven-inch knife that the Evolv scanner failed to detect at a school.

Financial

Reported: Yes, the report describes significant financial expenditures by school districts on the subscription-based system.

Directly caused: Charlotte-Mecklenburg Schools paid 4.9 million dollars for a four-year high school subscription and was charged 2,100 dollars for a replacement tablet.

Indirectly caused: Potential waste of public funds on ineffective security systems, including a planned 9.9 million dollars middle school expansion.

Inferred additional harm: Other school districts among the 800+ schools using Evolv likely spent millions of dollars on these ineffective systems.

Civil rights

Reported: Yes, the reports mention students losing privacy when entering schools using Evolv systems.

Indirectly caused: Students' privacy rights were compromised through invasive scanning and subsequent manual searches without a corresponding increase in safety.

Privacy

Reported: Yes, the reports explicitly discuss students losing privacy when entering schools using Evolv systems.

Indirectly caused: The deployment of the scanners compromised student privacy by subjecting them to automated physical scans and subsequent manual bag searches due to high false positive rates.

Inferred additional harm: Hundreds of thousands of students across 800+ schools had their privacy expectations compromised.

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 141000

Potential causes

Management

• Prioritizing Sales Over Safety: Evolv marketed unproven AI scanners to exploit school safety anxieties.
• Suppression of Product Flaws: Evolv hid instruction manuals and accused critics of endangering children.
• Ignoring Critical Test Results: School boards ignored warnings that default settings failed to detect Glocks.

Technology

• Flawed Classification Algorithm: AI failed to distinguish weapons from everyday items like school binders.
• Sensitivity Calibration Tradeoff: Lower settings missed Glocks while higher settings caused extreme false alarms.
• Brittle Hardware and Software: Scanners suffered frequent error messages and required manual updates.

Data Inputs

• Misclassified Everyday Objects: Laptops, binders, and umbrellas were misidentified as concealed weapons.

Human Factors

• Blind Trust in Vendor Claims: Administrators bought unproven AI systems without independent verification.
• Operator Distraction: Guards focused entirely on manual bag searches instead of hall monitoring.

Process and Methods

• Lack of Independent Testing: Districts failed to test if the scanners actually detected common handguns.
• Deceptive Marketing Practices: Vendor falsely claimed 70 percent labor reduction and perfect detection.
• Inefficient Maintenance Process: Software updates required tedious manual on-site programming by technicians.

Regulatory Environment

• Absence of AI Safety Standards: No federal standards verified AI weapons detection claims before deployment.
• Delayed Oversight Enforcement: FTC and SEC investigations occurred only after widespread school deployment.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide extensive, concrete evidence from FTC complaints, SEC investigations, and internal school district emails obtained via public records requests, detailing specific technical failures and operational impacts.
• Ambiguities identified: The exact sensitivity settings used during the stabbing incident are not fully disclosed due to security concerns.
• Alternative interpretations: None. The failure of the AI system to perform as advertised is well-documented by regulatory bodies and internal communications.

AI-powered weapons scanners deployed in US schools failed to detect actual weapons (resulting in a student stabbing) and generated high rates of false positives. While causing operational disruption and localized safety failures, the incident represents a minor national security threat, highlighting vulnerabilities in AI procurement and performance verification for public safety.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents an ongoing strategic concern regarding the verification of AI-powered security systems rather than an active, immediate crisis.
• Autonomy: Human-supervised. The AI operates autonomously to flag threats, but humans (school security staff) supervise the system and execute the final security actions.
• Novelty: Evolved capability. Represents an evolution of traditional metal detection technology into automated AI-driven screening, introducing new failure modes.

Impact by dimension

• Physical security: Minor. The AI's failure to detect a knife led to a student stabbing. While tragic, this remains a localized public safety issue manageable by local authorities, with minor national security impact.
• Information security: Negligible. No information warfare, deepfake operations, or compromise of classified intelligence capabilities occurred.
• Sovereignty: Negligible. Core sovereign government functions, electoral systems, and border controls were not impacted by the deployment of these commercial scanners in schools.
• Economic security: Negligible. While local school districts suffered financial losses due to ineffective procurement, there was no threat to national economic stability or strategic technological competitiveness.
• Societal stability: Minor. The deployment compromised student privacy and caused operational friction across multiple schools, but did not lead to large-scale civil unrest or systematic algorithmic oppression.