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

March 22, 2022

Evolv's AI-driven weapons detection systems, touted as superior alternatives to metal detectors, allegedly failed to accurately identify weapons while generating numerous false positives according to the FTC. This has led schools to trust an unreliable detection system, with at least one reported incident in October 2022 resulting in a missed knife and a student getting stabbed. It is crucial to ensure safe and secure AI practices, especially when it comes to safeguarding educational institutions. For those interested in shaping the future of responsible AI governance and harm prevention, consider joining the HISPI Project Cerebellum TAIM initiative by visiting JOIN US. By mapping incidents like this one to the HISPI Project Cerebellum TAIM (Measure), we can work together towards improving the safety and reliability of AI systems.

Matched TAIM controls

Suggested mapping from embedding similarity (not a formal assessment). Browse all TAIM controls

Alleged deployer
charlotte-mecklenburg-school-district
Alleged developer
evolv-technology
Alleged harmed parties
teachers-at-charlotte-mecklenburg-schools, students-at-charlotte-mecklenburg-schools, security-officers-at-charlotte-mecklenburg-schools, parents-of-students-at-charlotte-mecklenburg-schools, school-administrators-at-charlotte-mecklenburg-schools

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/349

Data source

Incident data is from the AI Incident Database (AIID).

When citing the database as a whole, please use:

McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.

Pre-print on arXiv · Database snapshots & citation guide

We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.