Las Vegas Police Are Running Lots Of Low Quality Images Through Their Facial Recognition System

from the that's-going-to-end-badly dept

Even when facial recognition software works well, it still performs pretty poorly. When algorithms aren't generating false positives, they're acting on the biases programmed into them, making it far more likely for minorities to be misidentified by the software.

The better the image quality, the better the search results. The use of a low-quality image pulled from a store security camera resulted in the arrest of the wrong person in Detroit, Michigan. The use of another image with the same software -- one that didn't show the distinctive arm tattoos of the non-perp hauled in by Detroit police -- resulted in another bogus arrest by the same department.

In both cases, the department swore the facial recognition software was only part of the equation. The software used by Michigan law enforcement warns investigators search results should not be used as sole probable cause for someone's arrest, but the additional steps taken by investigators (which were minimal) still didn't prevent the arrests from happening.

That's the same claim made by Las Vegas law enforcement: facial recognition search results are merely leads, rather than probable cause. As is the case everywhere law enforcement uses this tech, low-quality input images are common. Investigating crimes means utilizing security camera footage, which utilizes cameras far less powerful than the multi-megapixel cameras found on everyone's phones. The Las Vegas Metro Police Department relied on low-quality images for many of its facial recognition searches, documents obtained by Motherboard show.

In 2019, the LVMPD conducted 924 facial recognition searches using the system it purchased from Vigilant Solutions, according to data obtained by Motherboard through a public records request. Vigilant Solutions—which also leases its massive license plate reader database to federal agencies—was bought last year by Motorola Solutions for $445 million.

Of those searches, 471 were done using images the department deemed “suitable,” and they resulted in matches with at least one “likely positive candidate” 67% of the time. But 451 searches, nearly half, were run on “non-suitable” probe images. Those searches returned likely positive matches—which could mean anywhere from one to 20 or more mugshots, all with varying confidence scores assigned by the system—only 18% of the time.

Fortunately, low-quality images seemingly rarely return anything investigators can use. (Although that 18% is still 82 "likely positive matches...") If the system did, we'd be seeing far more bogus arrests than we've seen to this point. Of course, prosecutors and police aren't letting suspects know facial recognition software contributed to their arrests, so courtroom challenges have been pretty much nonexistent.

Although most of the information in the documents is redacted -- making it difficult to verify LVMPD claims about the software's contribution to arrests and prosecutions -- enough details remained to provide a suspect facing murder charges with information the LVMPD had never turned over to him or admitted to in court.

Clark Patrick, the Las Vegas attorney representing [Alexander] Buzz, told Motherboard that neither the LVMPD nor the Clark County District Attorney’s office ever informed him that investigators identified Buzz as a suspect using, at least in part, facial recognition technology. The Clark County District Attorney’s office did not respond to an interview request or written questions.

Had this information been given to Buzz and his attorney at the beginning of the trial, he likely would not have waived his right to a preliminary evidentiary hearing. If this had taken place -- along with knowledge of a private company's contribution to the investigation -- prosecutors may have had to produce information about the tech and the surveillance footage it pulled images from.

The documents don't appear to show a reliance on low-quality images to make arrests, but they do show investigators will run nearly any image through the software to see if it generates some hits. The precautions taken after this matter most. If investigators are only considering matches to be leads, it will head off most false arrests. But if investigators take shortcuts -- as appears to have happened in Detroit -- the outcome is disastrous for those falsely arrested. A person's rights and freedoms shouldn't be at the mercy of software that performs poorly even when given good images to work with. The use of this software is never going to go away completely, but agencies can mitigate the damage by refusing to treat matches as probable cause.

Filed Under: facial recognition, las vegas, law enforcement, police

NIST Study Confirms The Obvious: Face Masks Make Facial Recognition Tech Less Useful, More Inaccurate

from the for-now... dept

At the end of last year, the National Institute of Standards and Technology (NIST) released its review of 189 facial recognition algorithms submitted by 99 companies. The results were underwhelming. The tech law enforcement and security agencies seem to feel is a game changer is just more of the same bias we've been subjected to for years without any AI assistance.

Asian and African American people were up to 100 times more likely to be misidentified than white men, depending on the particular algorithm and type of search. Native Americans had the highest false-positive rate of all ethnicities, according to the study, which found that systems varied widely in their accuracy.

The faces of African American women were falsely identified more often in the kinds of searches used by police investigators where an image is compared to thousands or millions of others in hopes of identifying a suspect.

Who were the winners in NIST's facial recognition runoff? These guys:

Middle-aged white men generally benefited from the highest accuracy rates.

We have some good news and bad news to report from the NIST's latest facial recognition study [PDF]. And the good news is also kind of bad news. (The bad news contains no good news, though.)

The bad news is that the COVID-19 pandemic is still ongoing. This leads to the good news: face masks -- now a necessity and/or requirement in many places -- are capable of thwarting facial recognition systems.

Using unmasked images, the most accurate algorithms fail to authenticate a person about 0.3% of the time. Masked images raised even these top algorithms’ failure rate to about 5%, while many otherwise competent algorithms failed between 20% to 50% of the time.

But that's also bad news. This increases the chance of both false positives and false negatives. Both of these are unwelcome side effects of face coverings. The tiny bit of good news is that it generates mostly unusable images for passive systems (like those installed in the UK) that collect photos of everyone who passes by their lenses. The other small bit of good news in this bad news sandwich is this: face masks reduce the risk of bogus arrests/detainments.

While false negatives increased, false positives remained stable or modestly declined.

NIST also noticed a couple of other quirks in its study. Mask coverage obviously matters. The more that's covered, the less likely it is software will draw the correct conclusion. But color also matters. Black masks produced more bad results than blue masks.

Companies producing facial recognition tech (89 algorithms were tested by NIST for this project) aren't content to wait out the pandemic. Many are already working on algorithms that use fewer features to generate possible matches. This is also bad news. While the tech may be improving, working around masks by limiting the number of data points needed to make a match is just going to generate more false positives and negatives. But companies are already training their AI on face-masked photos, many of which are being harvested from public accounts on social media websites. Dystopia is here to stay. The pandemic has only accelerated its arrival.

Filed Under: face masks, facial recognition, nist, pandemic

Fan Uses AI Software To Lipread What Actors Really Said In TV Series Before Chinese Authorities Censored Them

from the I-saw-what-you-said-there dept

It's hardly news to Techdirt readers that China carries out censorship on a massive scale. What may be more surprising is that its censorship extends to even the most innocuous aspects of life. The ChinAI Newsletter, which provides translations by Jeff Ding of interesting texts from the world of Chinese AI, flags up one such case. It concerns a Chinese online TV series called "The Bad Kids". Here's how the site Sixth Tone describes it:

Since its first episodes were released on China's Netflix-like video platform iQiyi in mid-June, "The Bad Kids" has earned sweeping praise for its plot, cinematography, casting, dialogue, pacing, and soundtrack. It's also generated wide-ranging online discussion on human nature due to the psychology and complex motivations of its characters.

However, as the Sixth Tone article points out, the authorities required "a lot of changes" for the series to be approved. One fan of "The Bad Kids", Eury Chen, wanted to find out what exactly had been changed, and why that might be. In a blog post translated by ChinAI, Chen explained how he went about this:

Two days ago, I watched the TV series "The Bad Kids" in one go, and the plot was quite exciting. The disadvantage is that in order for the series to pass the review (of the National Radio and Television Administration), the edited sequences for episodes 11 and 12 were disrupted, even to the point that lines were modified, so that there are several places in the film where the actor's mouth movements and lines are not matched, which makes the plot confusing to people. Therefore, I tried to restore the modified lines through artificial intelligence technology, thereby restoring some of the original plot, which contained a darker truth.

The AI technology involved using Google's Facemesh package, which can track key "landmarks" on faces in images and videos. By analyzing the lip movements, it is possible to predict the sounds of a Chinese syllable. However, there is a particular problem that makes it hard to lipread Chinese using AI. There are many homophones in Chinese (similar sounds, different meanings). In order to get around this problem, Chen explored the possible sequences of Chinese characters to find the ones that best match the plot at that point. As his blog post (and the ChinAI translation) explains, this allowed him to work out why certain lines were blocked by the Chinese authorities -- turns out it was for totally petty reasons.

Perhaps more interesting than the details of this particular case, is the fact that it was possible to use AI to carry out most of the lipreading, leaving human knowledge to choose among the list of possible Chinese phrases. Most languages don't require that extra stage, since they rarely have the same number of homophones that Chinese does. Indeed, for English phrases, researchers already claimed in 2016 that their AI-based LipNet achieved "95.2% accuracy in sentence-level, overlapped speaker split task, outperforming experienced human lipreaders".

It's clear that we are fast approaching a situation where AI is able to lipread a video in any language. That is obviously a boon for the deaf or hard of hearing, but there's a serious downside. It means that soon all those millions of high-quality CCTV systems around the world will not only be able to use facial recognition software to work out who we are, but also run AI modules to lipread what we are saying.

Follow me @glynmoody on Twitter, Diaspora, or Mastodon.

Filed Under: ai, censorship, china, facial recognition, lip reading

Detroit PD Now Linked To Two Bogus Arrests Stemming From Facial Recognition False Positives

from the no-problem,-it's-just-rights-and-freedom dept

Late last month, the first known false arrest linked to facial recognition software was reported. But that first in AI police work now appears to be merely a repeat offender. There have been two bogus arrests linked to facial recognition false positive. And both bogus arrests were performed by the same law enforcement agency, the Detroit Police Department. Elisha Anderson of the Detroit Free Press has the details on the first blown call by the PD's software.

The high-profile case of a Black man wrongly arrested earlier this year wasn't the first misidentification linked to controversial facial recognition technology used by Detroit Police, the Free Press has learned.

Last year, a 25-year-old Detroit man was wrongly accused of a felony for supposedly reaching into a teacher’s vehicle, grabbing a cellphone and throwing it, cracking the screen and breaking the case.

Detroit Police used facial recognition technology in that investigation, too.

This man, Michael Oliver, was charged with larceny for the May 2019 incident he didn't actually participate in. The report by the Free Press contains photos of both the person caught on the phone's camera and Michael Oliver. They highlight one major problem with facial recognition software: even if one could be persuaded the two faces are a close match (and they don't appear to be), the recording used by investigators to search for a match showed the suspect's bare arms. The person committing the crime had no tattoos. Michael Oliver's arms are covered with tattoos, running from the top of his hands all the way up to his shoulders.

The facial recognition software delivered its mistake to investigators, who included this mismatch in the photos they presented to the person whose phone had been grabbed.

During the investigation, police captured an image from the cellphone video, sent it for facial recognition and the photo came back to Oliver, the police report said.

After Oliver was singled out, a picture of his face was included in a photo lineup of possible suspects that was presented to the teacher.

A second person, a student, was also captured in the video with the suspect. The officer in charge of the case testified he didn’t interview that person though he'd been given that student’s name.

Once again, the Detroit PD and local prosecutors are promising the thing that has already happened twice won't happen again. There are new processes in place, although it's unclear when those policies went into effect. Oliver was arrested late last year. The other bogus arrest occurred earlier this year. In both cases, reporters were assured by law enforcement spokespeople that things have changed.

Here's what official say is now in place, even though it's too little too late for two Black men arrested and charged for crimes they didn't commit. There are "stricter rules" in effect. Matches returned by the system are not probable cause for anything, but can only be used as "investigative leads." Supposedly, this software will now only be used to identify people wanted for violent felonies.

Prosecutors are doing things a bit differently, too. But it's a reaction, rather than a proactive effort. t's only now -- after two false arrests -- that the prosecutor's office is mandating review of all facial recognition evidence by the city's top prosecutor, Kym Worthy. Investigators must also produce corroborating evidence before seeking to arrest or charge someone based on a facial recognition match.

This seems unlikely to change anything. Outside of the limitation to violent crimes, both of these cases could have gone through the review process -- along with the limited corroborating evidence (in both cases, crime victims picked the AI's mismatches out of a lineup) -- and still resulted in the arrest of these two men. In both cases, investigators ended their investigations after this step, even though they were given the opportunity to interview other witnesses. If corroboration is nothing more than discovering humans are just as bad at identifying people as the PD's software is, the mistakes the PD claims will never happen again will keep on happening.

Filed Under: detroit, detroit police department, facial recognition, false arrest, michael oliver

Clearview Calls It Quits In Canada While Under Investigation By The Privacy Commissioner

from the come-back-and-take-your-beating,-you-cowards dept

Clearview AI -- the facial recognition service that gives all kinds of entities access to billions of face images scraped from the web -- is suddenly scaling back on its aggressive expansion plans. Once the plaything of billionaires, the unproven AI has been sold to retailers, fitness centers, police departments, and a handful of human rights violators.

Soon after its existence was exposed by the New York Times, Clearview AI proceeded to announce its plans to expand worldwide -- something it hoped to achieve while being sued multiple times in its home country. Canada was in its sights, but not so much any more.

In February, the Privacy Commissioner of Canada announced its office would be investigating Clearview AI and its still-not-independently-tested algorithm. This was triggered by numerous reports from journalists exposing how Clearview obtained its massive database of photos and how it was being used by government agencies and private entities.

The Privacy Commissioner was not impressed with Clearview's apparent disregard for privacy, saying it was specifically looking into reports that Clearview was collecting and using personal info without consent.

The investigation is still ongoing. But it appears Clearview is hoping to dodge the worst of it by removing itself from the Canadian market:

Clearview AI has advised Canadian privacy protection authorities that, in response to their joint investigation, it will cease offering its facial recognition services in Canada.

This step includes the indefinite suspension of Clearview AI’s contract with the RCMP, which was its last remaining client in Canada.

Clearview is making like a cop under investigation and retiring before it's concluded. But that's not going to save Clearview from findings that are pretty much guaranteed to be damning.

The investigation of Clearview by privacy protection authorities for Canada, Alberta, British Columbia and Quebec remains open. The authorities still plan to issue findings in this matter given the importance of the issue for the privacy rights of Canadians.

And the investigation Clearview can't dodge by saying "you can't fire me, I quit" will continue, looking into the company's slurping of Canadians' data as well as any attempts it makes to divest itself of contested Canadian data as it promises to pull out.

Good riddance, I say. And Canada says. Not directly, as of yet, but when the investigation concludes, I doubt the Privacy Commissioner will be offering Clearview a "we had you all wrong, bro." Clearview is the pox upon the humanity that is the facial recognition arms race. Weaponizing people's online postings against them using unproven AI is just pure evil.

Filed Under: canada, facial recognition, privacy
Companies: clearview, clearview ai

Boston The Latest City To Ban Facial Recognition Use By Government Agencies

from the returning-the-public's-faces-to-them dept

San Francisco led the way. Then the entire state of California followed suit. And on the other side of the country, a few smaller cities in Massachusetts did the same thing: banned facial recognition.

It just makes sense. The tech that's out there is as dangerous as it is unproven. Mostly known for its false positive rates, facial recognition software has shown it's capable of amplifying existing biases into actionable "intel" with the power to severely disrupt people's lives.

It's not just hypothetical. Just recently, a Michigan man became the first false positive arrested. He was detained for 30 hours based on a mismatch delivered by facial recognition software. Even companies that have been pitching facial recognition tech to law enforcement agencies have pulled back in recent weeks, refusing to become part of the problem… at least for the time being. One company that has done nothing but sell tech to cops has decided it won't be adding facial recognition to its near-ubiquitous body cameras.

With all of this going on, news of another facial recognition ban in a major city is no longer surprising. But it's still welcome news.

Boston has banned the use of facial surveillance technology in the city, becoming the second-largest community in the world to do so.

The city council unanimously voted on Wednesday to ban the use of the technology and prohibit any city official from obtaining facial surveillance by asking for it through third parties. The measure will now go to Mayor Marty Walsh with a veto-proof majority.

Veto-proof… and yet:

Walsh's office said he would review the ban.

The mayor's office has spelled "sign" incorrectly.

The discussion leading to the vote included the case of Robert Williams, who was arrested by Michigan police officers after their software said a screen-grab from a lo-res CCTV camera matched Williams' drivers license photo. Even local law enforcement reps -- who say they aren't currently using the tech -- expressed their reservations.

During a hearing earlier this month, Boston Police Commissioner William Gross said the current technology isn't reliable, and that it isn't used by the department.

"Until this technology is 100%, I'm not interested in it," he said.

“I didn’t forget that I'm African American and I can be misidentified as well," he added.

The ban [PDF] says it's illegal for any city agency or official to obtain facial recognition tech, either directly or through a third party. It also forbids agencies from employing contractors who plan to utilize the tech.

There aren't many loopholes or carve outs. Law enforcement will be able to use evidence derived from facial recognition tech in investigations so long as this evidence wasn't generated by any entity covered by the ban. Hopefully, this won't result in a bunch of requests being sent of out-of-state/federal agencies for secondhand searches.

The bill also includes a loophole for phones, tablets, computers, and social media services that use facial recognition software to verify identity, saving the city the trouble of having to rewrite the law to ensure employees can still use their personal and city-issued electronics -- something San Francisco failed to consider when it became the first city in the country to ban the tech.

The new law also gives people the right to sue if they think they were the target of unlawful facial recognition tech. And it makes anything derived from the banned tech illegal, making it impossible to further investigations, use in court, or otherwise exploit for the city government's gain.

It's a good bill and it's on the way to joining the handful of pioneering anti-surveillance creep laws on the books around the nation. Given the current state of citizen/law enforcement relations, there are certainly more bans on the way.

Filed Under: bans, boston, facial recognition, privacy

Detroit Police Chief Says Facial Recognition Software Involved In Bogus Arrest Is Wrong '96 Percent Of The Time'

from the good-to-have-that-all-out-in-the-open dept

The law enforcement agency involved with the first reported false arrest linked to facial recognition software is talking about its software. The Detroit Police Department -- acting on a facial recognition "match" handed to it by State Police investigators -- arrested resident Robert Williams for allegedly shoplifting watches from an upscale boutique.

Williams did not commit this robbery. He even had an alibi. But the investigators weren't interested in his answers or his questions. They had a lo-res screen grab from the store's CCTV camera -- one that software provided by DataWorks Plus said matched Williams' drivers license photo. Thirty hours later, Williams was cut loose by investigators, one of which said "I guess the computer screwed up" after rewatching the camera footage with Williams present.

The officers ignored the bold letters on top of the "match" delivered by the software. The writing said "This document is not a positive identification." It also said the non-match was not "probable cause for arrest." Unfortunately for the misidentified Michigan resident, the cops who arrested him treated the printout as both: positive identification and probable cause.

The policies governing law enforcement's use of this tech have changed since Williams' arrest in January. Under the current policy, lo-res images like the one that led to this arrest are no longer allowed to be submitted to the facial recognition system. That fixes a very small part of the problem. The larger problem is that the tech is mostly good at being bad. This isn't a complaint from critics. This comes directly from the top of the DPD.

In a public meeting Monday, Detroit Police Chief James Craig admitted that the technology, developed by a company called DataWorks Plus, almost never brings back a direct match and almost always misidentifies people.

“If we would use the software only [to identify subjects], we would not solve the case 95-97 percent of the time,” Craig said. “That’s if we relied totally on the software, which would be against our current policy … If we were just to use the technology by itself, to identify someone, I would say 96 percent of the time it would misidentify."

So, it's a bad idea to rely on the software and nothing else when attempting to identify criminal suspects. What happened to Robert Williams supposedly should never happen again… according to policy. But there's the question of why the PD still uses the software if it's not adding much value to the investigative process. It seems officers don't care much for the tech and yet it's still in use. The police chief thinks it's still worth keeping around, even if it has yet to show any positive results.

Craig and his colleague, Captain Ariq Tosqui, said that they want to continue using facial recognition because they say it can be a tool to assist investigators even if it doesn’t often lead to arrest. But even when someone isn’t falsely arrested, their misidentification through facial recognition can often lead to an investigator questioning them, which is an inconvenience at best and a potentially deadly situation at worst. According to Tosqui, the technology has been used on a total of 185 cases throughout the years. “The majority of the cases the detective reported back that [the software] was not useful.”

What it can do, apparently, is add another layer of surveillance on top of what's already in place. And it will do what most law enforcement surveillance creep does: put more eyes on Black residents. Part of the mild reforms Detroit enacted -- rather than implement a facial recognition ban -- mandate the publication [PDF] of facial recognition data by the department. Since the beginning of this year, the PD has used the tech 70 times. All but two of those instances involved a Black resident's photos being submitted.

One of the main sources for uploaded photos is the Project Green Light (PGL) network of cameras [PDF]. These are installed in or around businesses that participate in the project. Participation comes at a cost: $1-6,000 for the initial investment and $1,600 a year in video storage fees. The 550 participants' cameras are supposedly monitored 24/7 at the PD's "real time crime center." Incoming footage can be subjected to other surveillance tech, like facial recognition software and automatic license plate readers.

Here's the panopticon the city wanted:

The City of Detroit put forth a Request for Proposals for a contractor to work closely with the city, DPD, and Motorola (Company that help set up the RTCC) to set up a “turn-key” facial recognition system that would work with the already existing infrastructure of the RTCC. They specifically asked that the facial recognition work on at least 100 concurrent real-time video feeds, be integrated into the PGL system, and can be used by officers with a mobile app.

It got the software it needed from DataWorks: a blend of two algorithms known as "Face Plus." It's scary stuff.

Face Plus is capable of automatically searching all faces that enter camera frames against photos in the entity’s database, alerting authorities to any algorithmic matches. Additionally, there is a “watchlist” option where persons of interest can be monitored and alerted for…

According to DataWorks Plus, in 2017, this repository contained 8 million criminal pictures and 32 million “DMV” pictures. As the Free Press reported in March 2019, almost every Michigan resident has a photo of them in this system.

This is the system the police chief calls 96% inaccurate and DPD investigators call useless. Perhaps this false arrest, which has made national news, will put the brakes on the planned expansion of PGL and Face Plus to all public transit stops and vehicles. Or maybe it will convince the city it's paying too much for something that barely works and take facial recognition away from the PD until something better -- and less biased -- comes along.

Filed Under: detroit, detroit pd, facial recognition, police, robert williams
Companies: dataworks plus

Facial Recognition Software Finally Gets Around To Getting An Innocent Person Arrested

from the great-job-everyone dept

Well, it's happened. The thing people have been warning about for years. A person lost some of their freedom due to a facial recognition mismatch. It may have only been 30 hours, but it should have been zero. And it might have been zero hours if investigators had bothered to read the disclaimers attached to its facial recognition search results.

According to the New York Times report, this is the first time a false positive has led to someone being arrested. Or, at least, the first time the public's been made aware of it. A few years ago, the FBI and a local law enforcement agency used "facial analysis" performed by humans to arrest the wrong man twice for two separate robberies. This time, it was software. And it took 30 hours away from an innocent person.

On a Thursday afternoon in January, Robert Julian-Borchak Williams was in his office at an automotive supply company when he got a call from the Detroit Police Department telling him to come to the station to be arrested. He thought at first that it was a prank.

An hour later, when he pulled into his driveway in a quiet subdivision in Farmington Hills, Mich., a police car pulled up behind, blocking him in. Two officers got out and handcuffed Mr. Williams on his front lawn, in front of his wife and two young daughters, who were distraught. The police wouldn’t say why he was being arrested, only showing him a piece of paper with his photo and the words “felony warrant” and “larceny.”

His wife, Melissa, asked where he was being taken. “Google it,” she recalls an officer replying.

Googling it would not have helped. Williams was taken away by cops and held for 30 hours, accused of shoplifting watches from an upscale store nearly two years earlier. All the police had to work with were a blurry, lo-res screengrab from the store's CCTV camera and facial recognition software provided by DataWorks Plus. DataWorks tests AI algorithms created by contractors by running searches using low-quality images. DataWorks does not provide any measurements of these algorithms' accuracy, however. It apparently packages up its collection of algorithms and sells access to government agencies.

Five months after the crime was committed, Michigan State Police digital image examiner Jennifer Coulson uploaded the image captured by the store's camera. But investigators apparently ignored the big, bold warning attached to the top of the search results.

After Ms. Coulson, of the state police, ran her search of the probe image, the system would have provided a row of results generated by NEC and a row from Rank One, along with confidence scores. Mr. Williams’s driver’s license photo was among the matches. Ms. Coulson sent it to the Detroit police as an “Investigative Lead Report.”

“This document is not a positive identification,” the file says in bold capital letters at the top. “It is an investigative lead only and is not probable cause for arrest.”

The State Police still maintain investigators did nothing wrong. They didn't only rely on the facial recognition search results. They also showed Williams' drivers license photo to the loss prevention person at the store (as part of a "six-pack" of facial photos) and she supposedly identified him.

The contractors that supplied the AI system to the State Police seem skeptical of everything that happened here -- even their own software's participation in the arrest of Williams.

[Brendan] Klare, of Rank One, found fault with [the loss prevention person's] role in the process. “I am not sure if this qualifies them as an eyewitness, or gives their experience any more weight than other persons who may have viewed that same video after the fact,” he said. John Wise, a spokesman for NEC, said: “A match using facial recognition alone is not a means for positive identification.”

One of the officers who questioned Williams said -- after viewing the surveillance video again with Williams in the room -- that he "guessed the computer got it wrong." Even so, Williams was held for another several hours before being released on $1,000 bond. According to his first-person account of this incident, the first 18 hours of his detainment were spent in an overcrowded holding cell without any interaction with investigators or the officers who arrested him.

Two weeks later, prosecutors dropped the charges, but left themselves the option to refile the charges in the future. And the prosecutor's office appears to think that's a possibility. The spokesperson says there's another witness investigators are interviewing and cops might try to run Williams in again for a crime he didn't commit.

And Williams has an alibi. He posted a video to his Instagram account during his commute home from work. This occurred during the time the store was being robbed. But investigators never bothered to check. And this investigation -- which appears to have stemmed almost solely from a "match" generated by a lo-res CCTV screengrab -- resulted in an innocent man having his life severely disrupted.

Filed Under: arrests, detroit, facial recognition, robert williams, wrong guy

Amazon Senses US Isn't Currently Receptive To Its Cop Tech, Puts Rekognition On Mothballs For A Year

from the choosing-your-battles dept

Amazon has spent a few years courting cops with its proprietary facial recognition tech. Documents obtained by the ACLU in 2018 showed the supermassive company was seeking to get a cut of law enforcement's billions by giving them steep discounts on Rekognition, Amazon's branded surveillance tech.

Unfortunately for Americans who may be subjected to this tech, Amazon's brand of facial recognition remains unproven. Tests run by the ACLU managed to get several Congressional reps mistaken for criminals. Amazon responded to the false positives by claiming the ACLU didn't run it at the right accuracy setting. ACLU went with "80% confidence," which is something the tech allows. Amazon claimed only results obtained at the "95% confidence" setting mattered, even though the company does not tell law enforcement to only run Rekognition at that setting.

Amazon has taken a lot of criticism for its tech and its law enforcement hard-sell. It has also taken heat for Rekognition's apparent faults. But it has basically responded with silence. And it had a chance to prove its tech could actually recognize faces, but it chose to sit out NIST's study of more than 80 facial recognition algorithms. Rekognition's accuracy remains unproven.

At long last, Amazon has put its finger to the wind. And it has discovered the air is on fire. Giving cops more surveillance tech is about as popular as most cops at the moment. Rekognition may not be able to accurately recognize faces, but Amazon seems to have a handle on recognizing the obvious: now is not the time for more surveillance tech. From Amazon's blog:

We’re implementing a one-year moratorium on police use of Amazon’s facial recognition technology. We will continue to allow organizations like Thorn, the International Center for Missing and Exploited Children, and Marinus Analytics to use Amazon Rekognition to help rescue human trafficking victims and reunite missing children with their families.

There are a couple of caveats here: the moratorium does not say it applies to other government users that may not be strictly law enforcement agencies. And the moratorium is being put in place to give Congress time to act. Congress may not get anything done in the next year, which means Rekognition could be opened up for business with no meaningful changes.

Then there's this:

We’ve advocated that governments should put in place stronger regulations to govern the ethical use of facial recognition technology, and in recent days, Congress appears ready to take on this challenge. We hope this one-year moratorium might give Congress enough time to implement appropriate rules, and we stand ready to help if requested.

First off, Amazon has hardly been the most vocal advocate of stronger facial recognition tech regulations. Second, this sort of largesse should always be viewed skeptically. When a dominant player in the field calls for more regulation, it's usually with an eye on ensuring it remains dominant. Demanding Congress do something mainly means "do something that our competitors won't be able to comply with." Facial recognition is a crowded field. NIST's study examined 89 different algorithms. Getting Congress to clear out the field a bit is always helpful.

But there's hardly any downside there. We need fewer purveyors of surveillance tech, not more. If it eliminates a few Clearviews, we're probably all better off. But all the same, be wary of large companies calling for more regulation. It usually means they're looking to harm their competitors, rather than become more ethical themselves.

Filed Under: facial recognition, moratorium, police, privacy, regulations, rekognition
Companies: amazon