The Associated Press reports that a recent pilot program testing AI-powered weapons scanners in New York City subway stations resulted in over 100 false alerts, but no firearms were detected. Data from the New York Police Department, released Wednesday night, shows that the Evolv-manufactured scanners flagged a total of 118 false positives during nearly 3,000 screenings conducted at 20 subway stations. The program, introduced by Mayor Eric Adams as part of a broader initiative to increase transit safety, has prompted debate over its accuracy and constitutionality.
Mayor Adams, a Democrat and self-described technology advocate, announced the pilot earlier this year, citing concerns over subway safety following two high-profile shootings in the transit system. The portable Evolv scanners, already in use at other city facilities, were placed at a limited number of subway entrances to detect potential weapons. Over a 30-day testing period, the NYPD performed 2,749 scans, flagging 118 passengers as potential threats—though none were found to be carrying firearms. The scanners did detect 12 knives, but police have not clarified whether these were prohibited blades or permissible tools like pocket knives.
The trial has drawn skepticism from both subway riders and civil liberties organizations. Critics argue that scanning the subway’s millions of daily passengers, who enter through hundreds of different access points, would be impractical. There are also concerns about privacy and civil liberties, given the extensive monitoring that the full implementation of such technology would require. While subway crime remains rare, the high-profile shootings and other sporadic incidents have kept safety in the public spotlight.
Despite assurances from Mayor Adams that the pilot’s results would be made public, the statement released by the NYPD provided limited detail. It did not disclose how long each screening took, the number of officers required to staff the scanners, or how many passengers declined to participate. The announcement comes as Evolv, the Massachusetts-based company behind the scanners, faces legal and regulatory challenges. The company is under federal investigation by the Federal Trade Commission and the U.S. Securities and Exchange Commission over its marketing practices, and investors recently filed a class-action lawsuit accusing Evolv executives of overstating the reliability of their detection technology.
The results of New York City’s scanner trial have further fueled criticism. The Legal Aid Society issued a statement Thursday describing the pilot as “objectively a failure,” suggesting that the data demonstrates the technology’s lack of reliability. “Given this failed pilot, all the other overwhelming evidence against using Evolv’s weapons detectors, and the surrounding controversies, including lawsuits and various investigations, we hope that this ill-conceived, fraught, and unwanted idea is finally shelved for good,” the statement read.
ew YoFor now, the city has yet to finalize any contract with Evolv, and an NYPD spokesperson recently confirmed that the department is still “evaluating the outcome of the pilot.” Mayor Adams, who has supported expanding Evolv’s scanners in other public spaces, such as schools, has not indicated whether the city plans to pursue further testing or implementation in the subway system.
Appears the gizmo is falsely accusing citizens and having them detained for interrogation…typical nutzie demoCrap.
VOTE TRUMP/VANCE 2024!
of course it’s possible no one was carrying.
Lol …..well such a system did correctly identify my concealed assault umbrella in my jacket pocket…….. missed everything else.
To carry of weapons by ‘Law Abiding’ on public transportation should be legal. Even if unarmed and a person defends themselves & others you might end up in prison, i.e. the Danial Penny trial.
The issue for the police is determining which ones are law abiding and which ones are not. The only solutions thought up so far are all or nothing–either let everyone carry or let none carry.
How much did this experiment cost in equipment, man hours and lack of coverage in high crime areas?
Essentially the intended purpose of the Patriot act.
Not part of the purchasing process. What is important is that “by Mayor Eric Adams” who skimmed 20% of the budget
Good for Mayor Adam’s, though I believe he forgot to include a kick up to the Big Guy which might be the cause of certain problems.
These public safety systems are either designed to give more power to the police surveillance state, or to funnel money to democrats – or both.
Any “AI powered” surveillance tool can easily be repurposed to look for MAGA hats, mullets, old white guys or any other person deemed an enemy of the state.
The surveillance state is already here, it is just not as sophisticated as the Chinese system with real time monitoring of everybody through facial recognition software. But it is coming, and AI will lead the way to “the promised land of a safe and secure society.”
Of course its, basically, a failure. I can’t find the link right now, I posted here it months ago, but even the company finally admitted, after close scrutiny of their claims, the failure rate in detecting actual guns was almost 96%. This AI based system is basically ‘snake oil’.
but on the subject of ‘artificial intelligence’ its self…
They call it ‘artificial intelligence’ but it really isn’t. Simply put: Its still all programmed ‘responses’ to ‘trigger’ (no pun intended) ‘conditions’.
Real AI is still a few hundred years away.
The lawsuit complaint alleges that defendants made false statements and/or concealed that: (1) Evolv materially overstated the efficacy of its products; (2) the lack of effectiveness of Evolv’s products with regard to detecting knives and guns led to an increased risk of undetected weapons entering locations such as schools; (3) Evolv deceived the general public, its customers, and its investors regarding the effectiveness of its products; and (4) as a result, defendants’ statements about its business, operations, and prospects, were materially false and misleading and/or lacked a reasonable basis at all relevant times.
The plaintiffs go on to allege Evolv had, “taken significant action to make it appear that its products were effective, including manipulating test results.”
Leave it to leftists to fall for another bait-and-switch scheme to sell them some good feelz…like the $41 billions that were lost by the UN’s green energy programs.
AI = Artificial Incompetence, Artificial Idiocy, Artificial Ignorance, et al, Never Intelligence.
The Government, Military, Industrials, Businesses, Educators & Politicians are all lapping it up so that they will have an inanimate object to blame all of their FUBARS on.
Not My Fault, the AI did it !!!
Let’s look at this from the perspective of good stewardship of limited financial resources.
New York City could spend what would undoubtedly be at least $100 million on these fancy-shmancy weapon detectors which apparently do not work. Or, about 1 out of every 8 New York City residents could purchase a modest handgun for a combined total of about $300 million and provide a far superior system for mitigating the cost (to individuals and society) of violent crime. And the proverbial frosting on the cake: for that combined total of about $300 million for 1 out of every 8 New York City residents to arm themselves, New York City would be able to mitigate the cost of violent crime EVERYWHERE in New York City, not just in the transit system.
Of course the above solution requires that people learn some new information and act on facts rather than feelings … which means the above solution will never come to pass. Somehow just about everyone seems to know that, which is why New York City was such a believable setting for the 1981 movie Escape from New York.
Carrying arms on public transportation IS LEGAL rules a ILLINOIS FEDRAL JUDGE.
But don’t try it till till the appeals are over.
I will reply with a citation soon.
Considering that the law prevents law abiding citizens who find it necessary to use public transportation, especially at night, defenseless, never made any rational sense. Even SCOTUS recognized as much in Bruen.
I recently saw a system that somehow uses the density of an object to alert of the possibility of a weapon. I saw firsthand that the screen being monitored will draw a box around the part of the body where the gun is carried. It didn’t draw a box around a small fixed blade knife.
It might have been TeraHertz based. Basically a devices “shines” radio waves that are on the verge of being infrared light (but still radio rather than optical in nature) and those will penetrate clothing yet reflect off of very dense and of course metallic objects. Then an imaging device sensitive to those radio waves creates an image of the person and whatever objects are on their person.
I guess we could describe it another way as being similar to radar which obviously penetrates clothing and yet bounces off of very dense (and of course metallic) objects. Then a detector creates an image of the reflected energy. Just as radar creates a picture of rainfall in the atmosphere, TeraHertz imaging creates a picture of dense objects on a person.
That is probably the underlying imaging system. Then artificial intelligence is supposed to “recognize” firearms and somehow alert someone. I suspect that the imaging is working properly and the artificial intelligence is not.
It depends, incoherent or coherent detection.
In an incoherent architecture, a THz beam energy illuminates the ‘object’ (e.g. gun, knife, other) and the back scattered THz signal energy is focused by the optics onto a THz detector array. This allows, on average, a gray-scale ‘object image’ of about 4 bit, at distances of ~3 meters to (in some cases) ~10 meters.
In a coherent architecture, the back scattered light energy is detected by superposing it with a reference beam energy that possesses a fixed frequency relation to the back scattered light energy. Such an architecture is currently not feasible for deployment, and has only worked under very tightly controlled lab type conditions.
Basically; The problem with coherent architecture is commercially available THz components are not mature enough to be integrated into reliable coherent imaging schemes. So, overall, while they ‘work’ in the lab under controlled conditions, they are not reliable enough for a positive detection rate use in the outside world because conditions in the world outside the lab vary where in the lab the conditions are tightly controlled and, basically, limited to just what will work with commercially available THz components.
So to ride a NY subway I must subject my body to being blasted by even more radio waves.
Coronavirus vaccines and radio waves, and we wonder why the worlds going crazy.