Documents Show Hundreds Of Cops Have Run Clearview Searches, Often Without Their Employers' Knowledge Or Permission

from the doing-stuff-just-because-no-one-said-not-to dept

An impressive trove of public records obtained by BuzzFeed shows just how pervasive facial recognition tech is. Law enforcement agencies are embracing the tech, often with a minimum of accountability or oversight. That's how toxic tech purveyors like Clearview -- whose software relies on a multi-billion image database scraped from the web -- get their foot in the door to secure government contracts.

Despite being used for years, facial recognition tech has yet to prove it's capable of recognizing the right faces more often than the wrong ones. The accuracy gets even worse when it's deployed to recognize faces of women and minorities -- and given law enforcement's history of disproportionate enforcement -- it will be minorities harmed by the inaccurate tech more often than not.

What BuzzFeed has done with these Clearview records is compile a searchable database that allows readers to see if their local agencies have tried out the tech. Clearview's tech has yet to be subjected to outside review and its method of obtaining images -- scraping them from public posts on the web -- leaves a lot to be desired in terms of accuracy. (Unfortunately, as the EFF's Dave Maas points out, this doesn't mean BuzzFeed has made the dataset public -- only its interpretation of the data. But we'll take what we can get.)

The upshot? Lots and lots of experimentation. The downside? Very little oversight or explicit permission. According to the information BuzzFeed obtained, more than 335 US law enforcement agencies have at least tried out Clearview's facial recognition AI, and many of those searches had nothing to do with investigations.

Several of the responding agencies appear to be paying little attention to the actions of their employees:

Officials at 34 of those organizations said they were unaware that their employees had signed up for free trials until our questions prompted them to look.

Meanwhile, others pretended they had no responsive documents until asked twice:

Officials at another 69 entities at first denied their employees had used Clearview but later determined that some of them had.

While a tally of 335 law enforcement agencies may seem minute in comparison to the total number of law enforcement agencies in the United States, it would be wise to remember this dataset is far from complete. Nearly 100 agencies refused to answer definitively whether or not their employees had explored Clearview's offerings. Nearly 1,200 agencies have yet to turn over requested documents.

Meanwhile, Clearview has been out there touting law enforcement successes few law enforcement agencies will acknowledge. It also encourages the perception it is currently partnering with hundreds of cop shops while trying to divert attention away from its willingness to grant access to anyone interested in its unproven tech, including government agencies in countries known for their human rights abuses.

This data shows there's more interest in Clearview than law enforcement (as a whole) is willing to admit publicly. It also shows cops are playing with unproven tech -- quite possibly using images of people who aren't suspected of anything -- without the knowledge of their supervisors or government officials charged with overseeing their actions. What's left unsaid -- or unresponded to -- is at least as concerning as what has been admitted publicly. There's a rogue AI on the loose that links cops to a database filled with billions of images scraped from websites without their -- or their end users' -- approval.

Clearview is entirely problematic. But these agencies' willingness to exploit and examine this tech is even more so, considering the damage to rights and civil liberties the careless use of unproven facial recognition tech can cause.

Filed Under: facial recognition, police, privacy
Companies: clearview

Canadian Privacy Commission Says Clearview's App Is Illegal, Tells It To Pack Its Things And Leave

from the pics-and-GTFO dept

Clearview has screwed with the wrong people. The reprehensible facial recognition AI company that sells access to its database of scraped photos and personal info managed to raise the ire of some of the most restrained and polite people in the world, as Kashmir Hill reports for the New York Times.

The facial recognition app Clearview AI is not welcome in Canada and the company that developed it should delete Canadians’ faces from its database, the country’s privacy commissioner said on Wednesday.

“What Clearview does is mass surveillance, and it is illegal,” Commissioner Daniel Therrien said at a news conference. He forcefully denounced the company as putting all of society “continually in a police lineup.”

Clearview does appear to violate Canadian privacy laws, which require consent before using personal data. This was the impetus for a yearlong investigation of Clearview by Canadian privacy commissioners. The company claimed its offering was legal because it only utilized publicly available data scraped from dozens of social media sites. The commission disagreed.

“Information collected from public websites, such as social media or professional profiles, and then used for an unrelated purpose, does not fall under the ‘publicly available’ exception,” according to the report.

Clearview is going to court over this determination, saying it does nothing Google doesn't do and yet Google is still allowed to operate in Canada. Fair point, I guess, but Google doesn't appear to be selling government agencies access to billions of pieces of personal info for them to paw through at their leisure.

And Canadian law enforcement agencies are using Clearview to do just that. The commission noted that "thousands of searches" have been performed by dozens of agencies, including the Royal Canadian Mounted Police. Of course, this doesn't appear to be making Clearview much money. Only one agency actually paid for access. The rest of the "thousands" of searches were performed using trial accounts -- the ones Clearview encourages to "go wild" testing the AI by running it on pictures of friends, families, and anyone else they've got a picture of.

Clearview has already stopped selling access to the Canadian market, but that's not going to fix things. Clearview can control who it provides access to but it's going to have a much more difficult time determining who's in its database. If the illegality finding holds up, Clearview will need to delete information pertaining to Canadian residents. Finding Canadians in a database with billions of data points is something Clearview's AI can't handle, especially since it's much more difficult to determine whose information belongs to who when you've thrown it all into a big pile that's only expected to deliver matches to uploaded photos.

In the meantime, Clearview is offering Canadians the chance to opt out. All they have to do to be removed from Clearview's database is provide the company with personal information it may not have already collected. And then Canadians are asked to trust a company that's acted extremely carelessly and obnoxiously to follow through with its end of the bargain, rather than just add this new information to its existing stash.

For now, Clearview is merely facing the anger of Canadian regulators. There's really no legal force behind the commission's damnation. Not at the moment. But it could change in the future and it would make sense for Clearview to walk away from the mostly untapped Canadian market before it generates laws and legal precedent that would act as a blueprint for bans/removals in other countries.

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

Appeals Court Rejects Clearview's Attempt To Dodge A State Lawsuit By Trying To Make It A Federal Case

from the yeah...-it's-a-little-weird dept

Clearview's attempt to dodge a potential class-action lawsuit filed against it in Illinois has just been booted back to the Illinois court system by the Seventh Circuit Court of Appeals.

Clearview -- facial recognition's current supervillain -- was sued in Illinois by Illinoisans alleging violation of Illinois law. Multiple times.

The plaintiffs claimed Clearview's scraping of publicly available photos, location data, and other information from a variety of websites and social media platforms violated the state's law, which requires companies to obtain permission from people before harvesting and selling access to this data.

This same law netted a $550 million settlement from Facebook for its preemptive tagging of people in photos, something a court found to be a violation the law passed by the state in 2008. This settlement appeared to rattle Clearview, which filed documents with the court stating it would no longer do business in Illinois or knowingly collect biometric information from Illinois residents.

This lawsuit continues, however, thanks to the Seventh Circuit. In a somewhat novel move, the plaintiffs argued they do not have standing to pursue this lawsuit in federal court. Clearview argued otherwise, hoping to establish enough standing to take the lawsuit federal, at which point it would agree the plaintiffs did not have enough standing to move the case forward.

It's not often plaintiffs argue against their own standing, but the plaintiffs want this case in a state court, where they can pursue Clearview for violation of state laws. The only federal hook is Clearview's existence as a Delaware corporation headquartered in New York. State courts will normally allow lawsuits like these to be moved to federal court because the plaintiffs and defendant aren't located in the same state.

The district court, however, agreed with the plaintiffs: they did not allege any federal harms or anything else that would make the case better served at the federal level. The opening of the Seventh Circuit's opinion [PDF] highlights the bizarre nature of this appeal:

Oddly, [plaintiff Melissa] Thornley insists that she lacks standing, and it is the defendant, Clearview AI, Inc., that is championing her right to sue in federal court. That peculiar line-up exists for reasons that only a civil procedure buff could love: the case started out in an Illinois state court, but Clearview removed it to federal court. Thornley wants to return to state court to litigate the BIPA claims, but Clearview prefers a federal forum. The case may stay in federal court, however, only if the more stringent federal standards for standing can be satisfied; Illinois (as is its right) has a more liberal attitude toward the kinds of cases its courts are authorized to entertain. The district court held that Thornley has alleged only a bare statutory violation, not the kind of concrete and particularized harm that would support standing, and thus ordered the action remanded to the state court.

The court notes this reversal of roles has led to an equally unusual appeal. Normally, plaintiffs argue in favor of their standing to sue. In this case, the plaintiffs are arguing against this because it runs contrary to their interests and in favor of Clearview's. After examining far more precedent than one would imagine would be on hand, the Appeals Court says the case must exit the federal court system and return to Illinois.

The Appeals Court says the plaintiffs are free to craft their allegations as narrowly as they'd like. And as much as Clearview would like to be accused of even greater violations to (briefly) take the case federal for an easier dismissal, it's not up to the courts to tell plaintiffs how to craft their arguments or shuffle things from state to federal just because something a federal court could handle could plausibly be alleged.

[Thornley] does not contest either the existence of minimal diversity (she is a citizen of Illinois, and Clearview is a citizen of Delaware and New York) or the fact that more than $5,000,000 is at stake. Instead, she has simply offered a class definition that is narrower than it might have been. We have no reason to believe that the district court, acting on its own initiative, would certify a different and broader class; to that extent, the rule that the plaintiff controls her own case applies. And unlike the situation in Standard Fire, people who fall outside Thornley’s class definition are totally unaffected by this litigation. If they wish to sue Clearview, either alone or under a class definition that includes an allegation of injury, they are free to do so. Indeed, as we noted earlier, there are a number of class actions pending against Clearview, many of which appear to be broader than this one. We know of nothing that would prevent a putative class representative from taking a conservative approach to class definition. And if the plaintiffs change their tune in the state court, Clearview will be able to attempt to remove again to federal court, though we do not predict the outcome of such an effort.

Clearview's counterintuitive attempt to make a federal mountain out of the plaintiff's state molehill fails. The case remains in the state court system and limited to a specific violation of state law. Clearview will still have to defend itself against these claims.

Filed Under: facial recognition, illinois, jurisdiction, privacy, removal, standing
Companies: clearview

Facial Recognition Helps New Jersey Cops Jail The Wrong Man For Ten Days [Update]

from the bake-him-away,-toys dept

Update: The original report from NJ.com and the man's lawyer said that Clearview was involved, however the NY Times has now reported that it was not actually Clearview, but other facial recognition technology. The post has been updated accordingly.

Given facial recognition tech's predilection for false positives, the only thing surprising about its link to false arrests is how long it took. After years of live trial runs all over the world, tech used by the Detroit Police Department managed to contribute to two bogus arrests in a matter of months.

Despite this, the PD defended its use of the tech, even while admitting it was wrong 96% of the time. The PD said safeguards were in place to prevent false arrests -- including forbidding officers from using facial recognition search results as the sole probable cause for arrest warrants. Nonetheless, the false arrests happened. And, in both cases, it appears the safeguards were ignored.

Now, there's a third victim of facial recognition tech. A New Jersey man was falsely arrested for theft he didn't commit, thanks to mistaken facial recognition.

[W]hen police last year filed numerous charges against [NiJeer] Parks stemming from a shoplifting incident at a Woodbridge hotel in which the suspect hit a police car before fleeing the scene, the ex-convict who had worked eagerly to repair his life, tried just as hard to clear his name.

“I had no idea what this was about. I’d never been to Woodbridge before, didn’t even know for sure where it was,” Parks told NJ Advance Media last week.

On top of that, Parks didn't even own a car, much less drive one. He had to ask his cousin to drive him to the Woodbridge Police state after learning there was a warrant for his arrest. After that, he spent 10 days in jail before being freed -- a process that ended up costing him his savings and a year of his life just to clear his name.

So, how did the Woodbridge PD decide Parks was the suspect they were seeking? Investigators utilized facial recognition to "identify" the suspect in a January 2019 shoplifting of candy and snacks from a local hotel. (Of course, when cops argue they should be allowed to buy and use faulty surveillance tech, they always claim they'll only use it to track dangerous criminals, find kidnapping victims, and prevent child sexual abuse. In reality, tech gets used for things like this.)

Investigators fed photos of the suspect to the facial recognition database -- including the fake drivers license handed to cops by the suspect when officers confronted him at the scene. (The suspect escaped after a short car chase.) In an assertion no one should willingly associate with facial recognition, investigators called up the Woodbridge police to tell the PD they had a "high profile match." The "match" decided Parks was the suspect and things got much worse from there.

Parks voluntarily went to the station to "clear his name." Instead, he was handcuffed and interrogated for hours. Due to his prior convictions and the severity of the alleged crime (the suspect allegedly almost hit an officer while fleeing the scene), prosecutors "offered" six years in jail to Parks with the threat of a 20-year sentence if he took the case to trial.

Fortunately, the judge handling the case refused to consider the questionable facial recognition search results as evidence Parks had committed these crimes. Prosecutors dropped the case shortly after that, but Parks still spent ten days in jail, along with his life savings.

Now, Parks is suing police and prosecutors for attempting to ruin his life using nothing but sketchy technology that has since been banned for use by New Jersey law enforcement agencies. Hopefully, he'll be able to recover some of what he lost. And it adds to the body of evidence showing facial recognition shouldn't be trusted.

While the original reporting said that the facial recognition technology used by the police in this case came from Clearview, the NY Times has since said it appears to be existing government databases instead:

Mr. Parks’s mistaken arrest was first reported by NJ Advance Media, which said the facial recognition app Clearview AI had been used in the case, based on a claim in Mr. Parks’s lawsuit. His lawyer, Daniel Sexton, said he had inferred that Clearview AI was used, given media reports about facial recognition in New Jersey, but now believes he was mistaken.

[....]

According to the police report, the match in this case was to a license photo, which would reside in a government database, to which Clearview AI does not have access. The law enforcement involved in making the match — the New York State Intelligence Center, New Jersey’s Regional Operations Intelligence Center and two state investigators — did not respond to inquiries about which facial recognition system was used.

Filed Under: facial recognition, new jersey, nijeer parks, wrongful arrest
Companies: clearview, clearview ai

After Being Notified Of Info It Should Have Already Been Aware Of, LAPD Bans Clearview Use By Investigators

from the ohhhhhhh-those-unauthorized-uses dept

The Los Angeles Police Department is shutting down a very small percentage of its facial recognition searches. Last month, public records exposed the fact that the LAPD had been lying about its facial recognition use for years. Up until 2019, the department maintained it did not use the tech. Records obtained by the Los Angeles Times showed it had actually used it 30,000 times over the past decade.

The most recent development in the LAPD's mostly dishonest use of this tech is that it will not allow personnel to mess around with certain third-party offerings. As Buzzfeed reports, the LAPD has forbidden the use of Clearview by officers following the release of information the department already should have already been aware of.

Documents reviewed by BuzzFeed News showed that more than 25 LAPD employees ran nearly 475 searches with Clearview AI over a three-month period beginning at the end of 2019.

[...]

LAPD officials confirmed that investigators were using Clearview AI but declined to say which officers and which specific cases it was used for. They also refused to say whether the facial recognition software has led to arrests of any suspects.

Now that the public knows what the LAPD already should have known, the department is changing its policy to exclude Clearview… and probably not much else.

The Los Angeles Police Department has banned the use of commercial facial recognition systems, following inquiries from BuzzFeed News about its officers' use of a controversial software known as Clearview AI.

Have to love the fact that the LAPD needed to be apprised of what its own investigators were doing by journalists. That's the level of internal oversight we've come to expect from the nation's law enforcement agencies. If you don't look for anything, it's almost impossible to find misconduct and abuses of power. No news is the best news. And it can easily be achieved by doing nothing at all.

This ban only affects "commercial" software which means investigators will still be able to use (and misuse) more official products, like the facial recognition system owned by the county -- the same one the LAPD spent years denying it used.

And, although it's an incremental change that seems to only forbid the use of one particular facial recognition product, it's still good to see another law enforcement agency kick Clearview to the curb. Clearview's unproven AI trawls a database of photos scraped from the internet, making it a highly questionable addition to any government agency's surveillance repertoire. And Clearview has been highly irresponsible in its marketing and distribution of its tech, making unverified claims about law enforcement successes while encouraging government employees to test drive the software by feeding it faces of friends, family members, celebrities, etc.

If more agencies uninvite this third-party interloper, law enforcement critical mass will make Clearview's business plan untenable. It's already ditched most of its private customers in response to lawsuits. If the potential customers it has left refuse to do business with it, it will soon become little more than a horrible memory.

Filed Under: facial recognition, lapd, surveillance
Companies: clearview, clearview ai

Seattle PD Detective Took Clearview Facial Recognition Tech For A Spin, Possibly Violating Local Laws

from the when-any-hit-is-considered-a-success,-no-matter-how-many-misses-were-generated dept

It looks like some members of the Seattle Police Department have taken an interest in Clearview. Clearview scrapes photos and data from the open web and sells access to its untested facial recognition AI to government agencies, private companies, and the odd billionaire. According to Clearview, it has 4 billion scraped records in its database. What it doesn't have is a proven law enforcement track record for solving crimes, despite making extremely forward overtures to hundreds of law enforcement agencies around the globe.

Records [PDF] obtained by Bridget Brulolo of the Bridge Burner Collective show at least one Seattle PD investigator obtained access to this software and tried it out. This off-the-books test run may have broken a local law.

According to records recently released by the Seattle Police Department (SPD), a detective working for SPD signed up for and used facial recognition app Clearview AI, which appears to be in violation of the City of Seattle’s Surveillance Ordinance.

[...]

Released emails indicate that Nick Kartes, a detective on the South Precinct Burglary unit according to a staff roster obtained this summer, signed up for Clearview in September 2019 using his “@seattle.gov” work email address.

The emails show Nick Kartes took the software for a spin at least 30 times, using Seattle government computers to do so. He also downloaded the app to his phone and claimed to have "done some experimenting with it." It's unclear whether these "experiments" were related to Kartes' burglary investigations, but it must be said Clearview encourages government employees to "experiment" with the software using friends, family members, and random individuals as unwitting test subjects.

In fact, we'll just let Clearview say it, because this encouragement to "experiment" is included in one of the emails. [Emphasis in the original.]

Here are three important tips for using Clearview:

1. Search a lot. Your Clearview account has unlimited searches. Don't stop at one search. See if you can reach 100 searches. It's a numbers game. Our database is always expanding and you never know when a photo will turn up a lead. Take a selfie with Clearview or search a celebrity to see how powerful the technology can be.

This was followed up with another unproven assertion that encourages users to just feed photo after photo into the system.

Investigators who do 100+ Clearview searches have the best chances of successfully solving crimes with Clearview from our experience.

[...]

The more searches, the more matches. It's a numbers game. The investigators who search the most are the investigators that solve the most cases.

Yes, Clearview clearly advocates for responsible use of its tech to search its database of scraped images. Just feed faces into the portal or app and go wild. Fuck the consequences. Just imagine yourself solving crime after crime using this unproven AI. Bathe in the warm glow of being on the side of law and order as you prowl through the personal data of non-suspects.

Clearview's pitches sound like a mass mailing firm. "It's a numbers game." Given enough attempts, investigators are pretty much guaranteed to stumble across a usable lead. This ignores all the dead ends caused by false positives and false negatives. It also ignores the damage law enforcement can do with a handful of bad leads generated by "it's a numbers game" AI. Lives can be destroyed or severely disrupted. The innocent can be imprisoned. And all because a highly questionable facial recognition company that openly advocates for unethical use of its unethical scraped database says that's the best way to solve crimes.

Filed Under: facial recognition, nick kartes, seattle, seattle police department, surveillance
Companies: clearview, clearview ai

Clearview Resurfaces To Make Some Empty Promises About Keeping Cops From Abusing Its Tech

from the hey-maybe-knock-it-off-with-the-baseless-searches-guys dept

Belatedly realizing its reputation is a burning dumpster floating in a sewage retention pond, Clearview is finally trying to turn things around. Building a database of personal info and photos scraped from public websites, the company turned its product over to anyone who was interested. Private companies, billionaires, dozens of police departments -- all were invited to play with Clearview's toy, which set an untested facial recognition algorithm loose in the billions of images in its database.

Clearview likes to claim its tool is helping solve crimes. Police departments referred to in marketing materials and public statements have almost always disagreed with Clearview's self-assessment. Now, Clearview is trying to patch up its relationship with the public by altering its relationship with law enforcement agencies. CEO Hoan Ton-That is promising some sort of reform effort here, but his promises of a better, more trustworthy Clearview are as empty as its database is full.

Speaking at The Wall Street Journal’s Tech Live virtual conference, Mr. Ton-That said Clearview would make available training and compliance features that help ensure police officers use the technology ethically to solve crimes, though it would be up to police department heads to monitor officers and enforce the rules. One new feature he described is a requirement that police enter a specific case number and crime for each search, to enable better auditing.

This promise is coming from the same company that encouraged officers to "run wild" by testing the nascent AI on pictures of families, friends, and whoever else's photos they had access to. Now, this company is claiming it can train officers to use it "ethically" and provide "compliance features" (whatever those are) to ensure only good things are done with its database full of scraped personal info.

The backlash against facial recognition tech has pretty much cleared the playing field for Clearview. With Microsoft, Amazon, and IBM exiting the market at least temporarily, Clearview is poised for growth. But empty assurances that more "training" will be made available isn't going to make Clearview seem like a more trustworthy option. Neither is leaving it up to police departments to police their own use of the tech. If there's one thing law enforcement agencies have done consistently, it's prove they're incapable -- or at least unwilling -- to police themselves.

And Clearview is similarly unwilling to police itself. It may have pulled the plug on private customers as the result of multiple lawsuits, but it's still not very particular about who it sells to and what they'll do with the tech. If a government agency buys the software to combat child sexual abuse but then decides it might be similarly useful for hunting down undocumented immigrants, that's none of Clearview's business.

“It’s not our job to set the policy as a tech company,” he said. “It’s up to us to help them execute what they want to do…We are here to help the government and fulfill their vision in keeping us safe.”

If so, then Clearview feels it's not its job to rein in cops who might abuse the tech. That's why it's offering some undefined training and leaving the rest up to law enforcement agencies. Nothing about this will curb abuse. And it seems Clearview doesn't care whether its product is abused. Only that it sells.

Filed Under: facial recognition, police
Companies: clearview, clearview ai

Shitbirds Of A Feather Flock Together: ICE Signs $274,000 Contract With Clearview

from the voted-Most-Hateable-for-two-years-running! dept

ICE continues to not care what anyone thinks of it. Its tactics over the past few years have turned it into one of the federal government's most infamous monsters, thanks to its separation of families, caging of children, unfettered surveillance of undocumented immigrants, its fake university sting created to punish students trying to remain in the country legally, its sudden rescinding of COVID-related distance learning guidelines solely for the purpose of punishing students trying to remain in the country legally… well, you get the picture.

Perhaps it's fitting ICE is buying tech from a company that appears unconcerned that most of the public hates it. Clearview -- the facial recognition software that matches uploaded facial images with billions of images scraped from the open web -- is one of the latest additions to ICE's surveillance tech arsenal.

Immigration and Customs Enforcement (ICE) signed a contract with facial recognition company Clearview AI this week for “mission support,” government contracting records show (as first spotted by the tech accountability nonprofit Tech Inquiry). The purchase order for $224,000 describes “clearview licenses” and lists “ICE mission support dallas” as the contracting office.

That its new partner is being sued in multiple states (including a suit filed by the Vermont Attorney General) doesn't appear to concern ICE, which is itself often on the receiving end of lawsuits. Clearview may be making good on its promise to pull out of the private market and sell only to government agencies, but that just means it will be only governments using unproven AI and scraped images to pursue investigations and arrest people.

Clearview's statement to The Verge about its contract with ICE makes it appear this is all about the children:

“Clearview AI’s agreement is with Homeland Security Investigations (HSI), which uses our technology for their Child Exploitation Unit and ongoing criminal investigations,” Clearview AI CEO Hoan Ton-That said in an emailed statement to The Verge. “Clearview AI has enabled HSI to rescue children across the country from sexual abuse and exploitation.”

Clearview leans on the children here, but the statement says a couple of other interesting things I'm sure Ton-That would rather slide by unnoticed. First: "and ongoing criminal investigations." ICE considers every undocumented immigrant to be a criminal, which means the agency is going to use this software to track down people in the US for committing the civil violation of staying in the country without the proper paperwork. ICE has not been able to find enough dangerous immigrants to make the administration's public statements about rampaging hordes of bad hombres come true, so it has decided to go after everyone, including students here on visas.

Second, Clearview claims it has "enabled" HSI to "rescue children across the country." I'd say we'd just have to take its word on this but we certainly don't have to take its word on this because it's said things like this in the past only for the named law enforcement agency to contradict Clearview's claims when asked for comment. That may be what Clearview hopes its partnership with ICE/HSI will do, but it's difficult to believe the recently signed contract has already produced results or that ICE/HSI are really using this mostly to fight child sexual exploitation rather than just uploading photos of brown people and hoping for hits.

Either way, we can safely conclude both partners here suck. ICE is bad and keeps getting worse, and Clearview isn't ever going to improve and is presumably still scraping sites for "content" it can sell to its customers.

Filed Under: dhs, facial recognition, for the children, hsi, ics, surveillance
Companies: clearview, clearview ai