AI incident #831 ·
NYC Subway AI Weapons Scanners Yield High False Positive Rate and Detect No Guns in Month-Long Pilot Test
What happened
NYC implemented an AI enabled weapons scanner for a month-long pilot with limited success. Despite not finding any weapons during the September 2024 testing phase, there were 118 false positives in which a person was searched under suspicion of carrying a weapon with no actual gun detections.
Editor's notes (AI Incident Database)
The 118 false positives can be considered a privacy invasion, and according to some cited legal advocacy groups, a violation of due process. Reconstructing the timeline of events: (1) March 28, 2024: Mayor Eric Adams announces plans to deploy Evolv's AI-powered weapons scanners in selected NYC subway stations. (2) Summer 2024: Pilot program begins, deploying AI scanners across 20 subway stations. (3) September 2024: NYPD completes a 30-day testing period with the scanners, performing 2,749 scans and recording 118 false positives and no firearms detections. (4) October 23, 2024: NYPD releases a brief statement summarizing the pilot results, which is marked at the incident date for our purposes, even though each false positive (as well as the potential for firearms to have slipped past detection) may be considered discrete incidents in and of themselves.
Only the incident metadata is stored here. The underlying news reports are on the AI Incident Database (CC BY-SA 4.0); use the links above to read them.
News reports (2)
Coverage catalogued by the AI Incident Database. Titles link to the original publisher; the text is not reproduced here.
Who was involved
- Alleged deployer
- New York City Government Law enforcement
- Alleged developer
- Evolv Technology
- Alleged harmed party
- New York City subway riders
AI systems implicated
Classification (MIT AI Risk Repository taxonomy)
- Risk domain
- AI system safety, failures, and limitations
- Risk subdomain
- 7.3 Lack of capability or robustness
- Causal entity
- AI
- Intent
- Unintentional
- Timing
- Post-deployment
- Harm level
- —
- Sectors
- —
- Countries
- —
Risk entries describing this failure mode
Entries from the MIT AI Risk Repository coded to subdomain 7.3.
- Reliability issues
"Relying on general-purpose AI products that fail to fulfil their intended function can lead to harm. For example, general- purpose AI systems can make up facts (‘hallucination’), generate erroneous computer code, or pro...
- Type 2: Bigger than expected
Harm can result from AI that was not expected to have a large impact at all, such as a lab leak, a surprisingly addictive open-source product, or an unexpected repurposing of a research prototype.
- Type 3: Worse than expected
AI intended to have a large societal impact can turn out harmful by mistake, such as a popular product that creates problems and partially solves them only for its users.
- Ethics and Morality Issues
LMs need to pay more attention to universally accepted societal values at the level of ethics and morality, including the judgement of right and wrong, and its relationship with social norms and laws.
- Safe learning
"AGIs should avoid making fatal mistakes during the learning phase. Subproblems include safe exploration and distributional shift (DeepMind, OpenAI), and continual learning (Berkeley)."
- Malign belief distributions
"Christiano (2016) argues that the universal distribution M (Hutter, 2005; Solomonoff, 1964a,b, 1978) is malign. The argument is somewhat intricate, and is based on the idea that a hypothesis about the world often includ...
- Meta-cognition
"Agents that reason about their own computational resources and logically uncertain events can encounter strange paradoxes due to Godelian limitations (Fallenstein and Soares, 2015; Soares and Fallenstein, 2014, 2017) an...
- Misaligned consequentialist reasoning
"As we think about even more intelligent and advanced AI assistants, perhaps outperforming humans on many cognitive tasks, the question of how humans can successfully control such an assistant looms large. To achieve the...
Related incidents on the AI Incident Database
Linked by AIID editors or by its text-similarity model.
Incidents in the same risk subdomain
- Purported AI Name-Reading System Reportedly Skipped and Misannounced Graduates at Arizona's Glendale Community College Commencement
- PocketOS Production Database Was Reportedly Deleted by Cursor AI Agent Running Claude Opus 4.6
- Baidu Apollo Go Robotaxis Stopped in Traffic During Reported System Failure in Wuhan, Stranding Some Passengers
- Purportedly AI-Enabled Targeting System Was Reportedly Implicated in Deadly U.S. Strike on Iranian Primary School
- Claude Code Agent Reportedly Deleted DataTalks.Club Production Infrastructure, Database, and Snapshots via Terraform
- Purportedly AI-Generated Sepsis Alert Reportedly Prompted Potentially Inappropriate IV Fluid Administration for a Dialysis Patient, Averted by Clinician Intervention