Clearview Says Section 230 Immunizes It From Vermont's Lawsuit Over Alleged Privacy Violations

from the assuming-the-judge-will-call-this-theory-'novel'-which-isn't-a-compl dept

Clearview is currently being sued by the attorney general of Vermont for violating the privacy rights of the state's residents. As the AG's office pointed out in its lawsuit, users of social media services agree to many things when signing up, but the use of their photos and personal information as fodder for facial recognition software sold to government agencies and a variety of private companies isn't one of them.

[T]he term "publicly available" does not have any meaning in the manner used by Clearview, as even though a photograph is being displayed on a certain social media website, it is being displayed subject to all of the rights and agreements associated with the website, the law, and reasonable expectations. One of those expectations was not that someone would amass an enormous facial-recognition-fueled surveillance database, as the idea that this would be, permitted in the United States was, until recently, unthinkable.

Thus, when an individual uploads a photograph to Facebook for "public" viewing, they consent to a human being looking at the photograph on Facebook. They are not consenting to the mass collection of those photographs by an automated process that will then put those photographs into a facial recognition database. Such a use violates the terms under which the consumer uploaded the photograph, which the consumer reasonably expects will be enforced.

This is somewhat the same point multiple companies have made with their (ultimately ineffective) cease-and-desist orders: we have not agreed to allow Clearview to harvest data from our sites and sell that collected data to others.

Whether or not selling this scraped collection to law enforcement agencies is unlawful in Vermont remains to be seen. But Clearview is fighting back in court, raising a truly questionable Section 230 defense against the AG's lawsuit.

Clearview is represented by Tor Ekeland, who has been truly useful in defending people against bogus prosecutions. But Ekeland appears to believe Section 230 is a net loss for the public, so it's interesting to see him raise it as a defense here.

Clearview's motion to dismiss [PDF] compares Clearview to Google, claiming its bots crawl the web and cache images (and other data) on servers. However, Clearview claims it collects "far less data" than comparable search engines. According to its filing, Clearview does not collect any identifying info either -- at least not intentionally. It only harvests photos and their metadata. The company says only 10% of the photos in its 4-billion photo database have any metadata attached.

However, this doesn't mean the software can't compile a staggering amount of information on a person and return this long list in response to an uploaded facial photo. To comply with California data privacy laws, Clearview has given state residents the opportunity to see what Clearview has gathered on them. It's a lot.

The depth and variety of data that Clearview has gathered on me is staggering. My profile contains, for example, a story published about me in my alma mater’s alumni magazine from 2012, and a follow-up article published a year later.

It also includes a profile page from a Python coders’ meetup group that I had forgotten I belonged to, as well as a wide variety of posts from a personal blog my wife and I started just after getting married.

The profile contains the URL of my Facebook page, as well as the names of several people with connections to me, including my faculty advisor and a family member…

Clearview's assertions about the personal information it intentionally gathers are meant to head off the Vermont AG's claims that the company is violating state privacy laws. It's also meant to portray Clearview as no more damaging to privacy than search engines like Google and no more nefarious than a Google search. The problem is cops are less likely to trust a Google search and more likely to trust a company that says it has 4 billion images and 600 law enforcement "partners," even if the search results are equally questionable.

But on to the Section 230 argument, which is kind of amazing in its audacity.

Clearview is entitled to immunity under the CDA because: (1) Defendant is an interactive computer service provider or user; (2) Plaintiff’s claims are based on “information provided by another information content provider;” and (3) Plaintiff’s claims would treat Defendant as the “publisher or speaker” of such information.

These are the base claims. There's more to it. But this is a company raising a defense afforded to service providers who host third party content. Here, there is no third party content -- at least not in the sense that we're used to. The "third parties" Clearview deals with are government agencies, who contribute no content of their own and only search the database of scraped photos using uploaded images.

In addition, the Vermont AG is not seeking an injunction against Clearview because of any particular content in its database. For example, the lawsuit is not predicated on defamatory content a user created. Instead, it's suing Clearview because its method of database compilation ignores the state's privacy laws. It's hard to imagine how Section 230 fits this particular action, but this filing attempts to do exactly that.

First, Clearview asserts it's a search engine just like Google. And if Google can't be sued for violating privacy rights of users of other sites whose personal photos/information show up in Google searches, neither can Clearview.

The Attorney General seeks to prohibit Clearview from accessing or using publicly distributed photos, none of which Clearview AI created. Clearview AI's republication of third-party content is the result of its search engine algorithm, which in this instance happens to be a biometric facial algorithm. The underlying technology does not transform Clearview into an information content provider that would be ineligible for CDA immunity. The CDA protects the publication of search engine results.

[...]

Clearview's publication of its biometric facial algorithms results does not make it an information content provider any more than Google becomes one when it publishes its search algorithm results. Simply put, “[i]f a website displays content that is created entirely by third parties, ... [it] is immune from claims predicated on that content.”

If Clearview did not create the content, it cannot be held responsible for its use of it -- even if end users never specifically agreed to be part of a database accessible by law enforcement.

The motion also says Clearview cannot be viewed as a publisher, since all it has done is created (with scraped content) a searchable database of third-party content.

Vermont’s complaint cannot change the fact that it is targeting Clearview for performing the exact same functions as corporations like Google and Microsoft. Vermont claims that, “at a minimum, Clearview ‘must obtain the other party’s consent before’ using consumers’ photos from any website.” Google, in contrast, is said to “respect the Terms of Service of the websites they visit." But Google searches are filled with information that individuals wanted to remain private, such as nonconsensually distributed intimate images. Nevertheless, Google has repeatedly been protected by §230 because courts have correctly viewed it as a publisher.

[...]

Clearview’s use of a complex algorithm does not negate the fact that it is performing the traditional role of a publisher. The Second Circuit emphatically rejected a claim that Facebook’s “matching” algorithm deprecated its status as a publisher. The Second Circuit has stated that, “we find no basis in the ordinary meaning of ‘publisher,’ the other text of Section 230, or decisions interpreting Section 230, for concluding that an interactive computer service is not the ‘publisher’ of third-party information when it uses tools such as algorithms that are designed to match that information with a consumer's interests."

The state has responded [PDF] to Clearview's Section 230 assertions and has made the obvious point: this legal action isn't being brought over content generated by third parties that would normally be met with immunity arguments. It's being brought over Clearview's acquisition and use of third-party content that state residents never agreed to being harvested/used by the facial recognition tech company.

The injurious information that Clearview claims gives it Section 230 immunity are the photographs that it screen-scraped. But the photographs themselves, as they were posted on the internet by their owners, are not injurious, and do not give rise to any of the State’s claims. Put another way, the State’s cause of action is not properly against millions of individuals who posted anodyne photographs on the internet.

The State’s claims are not based on the specific information at issue. The photographs themselves do not give rise to any of these claims. The State’s claims are for unfairness, deception, and fraudulent acquisition of data. Compl. ¶¶ 76-86. Specifically, Clearview’s conduct in acquiring the photographs through fraudulent means (see Section VI infra, discussing use of term “fraudulent”), storing them without proper security, applying facial recognition technology in a manner meant to violate privacy and infringe civil rights, and providing access to the database to whomever wanted it without concern for the safety or rights of the public, give rise to the State’s claims.

The state's counter-argument hinges on a close reading of Section 230 -- one that turns on a certain voluntary action by third parties.

Section 230 requires that the information at issue be provided by the third-party content provider. Again, the common thread in Section 230 cases is that the Information Content Provider posted the offensive information on the defendant’s servers. Here, no Vermont consumer could have intentionally provided any photographs to Clearview’s servers, because prior to the discovery in January of this year that Clearview had 3 billion photographs in a New York Times exposé, the general public did not know that Clearview existed.

It also points out how Clearview differs from the search engines it tries to compare itself to favorably.

Clearview is not a search engine like Google or Bing. Clearview’s App does something that no other company operating in the United States, including search engine companies, has ever done. In fact, search engine companies that are capable of creating a product like Clearview’s refused to do so for ethical reasons.

The state isn't impressed by Clearview's arguments and sums everything up with this:

For the fact pattern to apply, the photographs themselves would have to somehow be unfair or deceptive and the State’s claims would more properly be brought against the individuals.

In essence, the lawsuit isn't about objectionable content hosted by Clearview, but objectionable actions by Clearview itself. That's why Section 230 doesn't apply. I'm not sure how the local court will read this, but it would seem readily apparent that Section 230 does not immunize Clearview in this case.

Filed Under: facial recognition, privacy, scanning, section 230, vermont
Companies: clearview, clearview ai

London's Facial Recognition Rollout Trips Over The Pandemic As Facemasks Render The System Even More Useless

from the of-course,-some-people-still-refuse-to-wear-them-so... dept

The rollout of London's facial recognition cameras -- hundreds of them -- is being held up by unexpected developments. No, it's not the efforts of legislators, privacy activists, or some sudden concern about the unreliability of the tech. It's the global pandemic, which has taken away many of the facial features the cameras are looking for. (h/t Jeffrey Nonken in the TD Chat window)

The Metropolitan Police is considering whether to pause its use of facial recognition cameras because so many Londoners are wearing face masks.

Use of the technology, which was deployed across London from January in a bid to spot serious criminals, is regarded as controversial by civil liberties campaigners.

One assembly member, Caroline Pidgeon, made the obvious point that a rollout during a pandemic necessitating face mask use isn't the best timing for tech that already has demonstrated an inability to correctly recognize unmasked faces.

This point is driven home in a letter composed by Pidgeon and Sian Berry, another assembly member. The letter says the tech's failures -- and the lack of public input or oversight -- means Londoners should only expect bad things from a large-scale rollout of live facial recognition.

In a letter published [PDF] on the London Assembly website, the two elected representatives said: "We both believe that the way in which LFR [live facial recognition] is being rolled out as an operational tool in London is ill-advised, and that this technology will have a chilling effect on civil liberties if it is not used with clarity, accountability and with full democratic consent."

The letter says the tech used by the Metropolitan Police isn't merely error-prone. It's pretty much all errors, all the time.

In a recent answer to a written question, the Mayor made clear that in the first two deployments of LFR this year more than 13,000 faces were scanned. Out of these 13,200 people scanned only six individuals were stopped. However, of those six, five were misidentified and incorrectly stopped by the police, not being wanted for any crime. Furthermore, of the total of eight people who created a ‘system alert’, seven were incorrectly identified. This is a concerning pattern of inaccuracy and misidentification of innocent people, which was also identified during the trials of this technology.

The letter also points out that London's law enforcement has refused to cooperate with the government's oversight of this tech, telling legislators to refer to previously published documents rather than directing them to anything more recent that might provide insight on the current state of the technology. The assembly members say the available documents do not fully address tests mandated by the London Policing Ethics Panel. They also note that while the Ethics Panel has suggested this tech only be used to investigate "serious crimes," the Metropolitan Police has made it clear it has no intention to comply with this guideline.

The letter asks more questions of the MPS which will likely also go unanswered. The assembly members want assurance these systems won't be deployed during protests, demonstrations, or public events. They also want to know if the forces of law and order are cool with blowing off Parliamentary directives.

The House of Commons Science and Technology Committee has called for a halt to the police's use of live facial recognition, until relevant regulations are in place. Do you feel comfortable ignoring Parliament?

The answer to the last one is probably "yes." The Metropolitan Police has continued to deploy the tech despite a constant stream of false positives. The only thing standing between it and a full rollout is the pandemic itself. The MPS can't honestly say its system will be a useful crime fighting tool when people's faces are partially hidden from view. The MPS couldn't honestly say this prior to en masse mask-wearing, but at least the system still had access to full faces. Continuing the rollout would show the police care more about playing with their toys then ensuring the safety and well-being of London residents.

Filed Under: cameras, covid-19, face masks, facial recognition, london, metropolitan police, privacy

In Response To Getting Sued, Clearview Is Dumping All Of Its Private Customers

from the thanks-for-the-half-assed-largesse,-asses dept

This is the first good news we've heard from Clearview since its exposure by Kashmir Hill for the New York Times back in January. In response to a lawsuit filed against it in Illinois accusing it of breaking that state's privacy laws with its scraping of images and personal info from a multitude of social media platforms, Clearview has announced it's cutting off some of its revenue stream.

Clearview AI — the controversial face-tracking company known for scraping more than 3 billion photos from social media sites including Facebook and Twitter — said it is ending its relationships with non–law enforcement entities and private companies amid regulatory scrutiny and several potential class action lawsuits.

Responding to one of those lawsuits, Clearview claimed in legal documents filed in an Illinois federal court on Wednesday that it was taking those voluntary actions and would “avoid transacting with non-governmental customers anywhere.”

Clearview's motion to dismiss [PDF] argues this lawsuit is now mooted since it will no longer be doing business with any Illinois entities and has apparently found some way to sort out personal info on Illinoians from the rest of scraped stash. Take a couple of grains of salt before reading the following:

Moreover, Clearview has taken steps to secure and implement limits regarding the retention of any Illinois photos. Clearview is terminating access rights to its app for all account holders based in Illinois and is terminating the accounts of any non-law enforcement or government entity.

This is better but it's still nothing great. Governments can still use a privacy law-violating app. That's still a big problem. While private companies may be using Clearview's app the most, it's the company's government customers that are more of a problem. Unproven tech may get you booted from the local mall for resembling a shoplifting suspect, but law enforcement agencies can ruin your life and take away several of your freedoms. And Clearview isn't exactly selective about who it sells to, so plenty of overtly abusive governments will still get to use untested facial recognition software to destroy lives without worrying about niceties like due process.

Presumably, this will end the use of Clearview's AI by Illinois government agencies. But reporters at BuzzFeed couldn't get any confirmation from the Chicago PD that its plug had been pulled.

While the Chicago Police Department did not return a request for comment, a spokesperson for the Illinois secretary of state confirmed the office had been using Clearview “for about six months” to “assist other law enforcement agencies.”

Law enforcement agencies make up most of Clearview's customers in the state. If Clearview's sworn statements are true, they should all find their accounts deleted in the very near future. But Clearview hasn't been too honest when dealing with judgments passed by the court of public opinion, so it's probably wise to hold our golf applause for this lawsuit-prompted pullout until more facts are in.

We also shouldn't read too much into its declaration that it will be restricting gathering data on Illinois residents. Clearview's one-state filtering system apparently relies on metadata -- a lot of which is stripped by sites when users upload photos. It also will avoid scraping data and images from sites with words like "Chicago" or "Illinois" in the URL, which will leave a lot of Illinois-based websites open for Clearview's scraping business, no matter what it may be asserting in court. Sure, this is Clearview "taking steps" to prevent violating state law, but these two specifics can hardly be considered ambulatory.

The better news is this:

In addition, the startup said it was implementing an “opt-out mechanism” to allow people to exclude photos from its database.

Again, no congratulations until we see this implemented, and only if there's some outside auditing done to ensure Clearview is actually doing the things it says it is. For that matter, the AI still hasn't been independently examined, so we don't even know what sort of false positive rate Clearview's four-billion-image database is capable of generating.

Clearview is still as sketchy as ever and only slightly less dangerous following this move. As long as it still sells its scraped database and unproven AI to governments, it's still a threat to millions of people.

Filed Under: facial recognition, governments, privacy, surveillance
Companies: clearview, clearview ai

Harrisburg University Researchers Claim Their 'Unbiased' Facial Recognition Software Can Identify Potential Criminals

from the fresh-hells-delivered-daily dept

Given all we know about facial recognition tech, it is literally jaw-dropping that anyone could make this claim… especially without being vetted independently.

A group of Harrisburg University professors and a PhD student have developed an automated computer facial recognition software capable of predicting whether someone is likely to be a criminal.

The software is able to predict if someone is a criminal with 80% accuracy and with no racial bias. The prediction is calculated solely based on a picture of their face.

There's a whole lot of "what even the fuck" in CBS 21's reprint of a press release, but let's start with the claim about "no racial bias." That's a lot to swallow when the underlying research hasn't been released yet. Let's see what the National Institute of Standards and Technology has to say on the subject. This is the result of the NIST's examination of 189 facial recognition AI programs -- all far more established than whatever it is Harrisburg researchers have cooked up.

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.

Why is this acceptable? The report inadvertently supplies the answer:

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

Yep. And guess who's making laws or running police departments or marketing AI to cops or telling people on Twitter not to break the law or etc. etc. etc.

To craft a terrible pun, the researchers' claim of "no racial bias" is absurd on its face. Per se stupid af to use legal terminology.

Moving on from that, there's the 80% accuracy, which is apparently good enough since it will only threaten the life and liberty of 20% of the people it's inflicted on. I guess if it's the FBI's gold standard, it's good enough for everyone.

Maybe this is just bad reporting. Maybe something got copy-pasted wrong from the spammed press release. Let's go to the source… one that somehow still doesn't include a link to any underlying research documents.

What does any of this mean? Are we ready to embrace a bit of pre-crime eugenics? Or is this just the most hamfisted phrasing Harrisburg researchers could come up with?

A group of Harrisburg University professors and a Ph.D. student have developed automated computer facial recognition software capable of predicting whether someone is likely going to be a criminal.

The most charitable interpretation of this statement is that the wrong-20%-of-the-time AI is going to be applied to the super-sketchy "predictive policing" field. Predictive policing -- a theory that says it's ok to treat people like criminals if they live and work in an area where criminals live -- is its own biased mess, relying on garbage data generated by biased policing to turn racist policing into an AI-blessed "work smarter not harder" LEO equivalent.

The question about "likely" is answered in the next paragraph, somewhat assuring readers the AI won't be applied to ultrasound images.

With 80 percent accuracy and with no racial bias, the software can predict if someone is a criminal based solely on a picture of their face. The software is intended to help law enforcement prevent crime.

There's a big difference between "going to be" and "is," and researchers using actual science should know better than to use both phrases to describe their AI efforts. One means scanning someone's face to determine whether they might eventually engage in criminal acts. The other means matching faces to images of known criminals. They are far from interchangeable terms.

If you think the above quotes are, at best, disjointed, brace yourself for this jargon-fest which clarifies nothing and suggests the AI itself wrote the pullquote:

“We already know machine learning techniques can outperform humans on a variety of tasks related to facial recognition and emotion detection,” Sadeghian said. “This research indicates just how powerful these tools are by showing they can extract minute features in an image that are highly predictive of criminality.”

"Minute features in an image that are highly predictive of criminality." And what, pray tell, are those "minute features?" Skin tone? "I AM A CRIMINAL IN THE MAKING" forehead tattoos? Bullshit on top of bullshit? Come on. This is word salad, but a salad pretending to be a law enforcement tool with actual utility. Nothing about this suggests Harrisburg has come up with anything better than the shitty "tools" already being inflicted on us by law enforcement's early adopters.

I wish we could dig deeper into this but we'll all have to wait until this excitable group of clueless researchers decide to publish their findings. According to this site, the research is being sealed inside a "research book," which means it will take a lot of money to actually prove this isn't any better than anything that's been offered before. This could be the next Clearview, but we won't know if it is until the research is published. If we're lucky, it will be before Harrisburg patents this awful product and starts selling it to all and sundry. Don't hold your breath.

Filed Under: ai, artificial intelligence, bias, facial recognition, precrime
Companies: harrisburg university

Hard Pass: Clearview Offers To Help Out With COVID-19 Contact Tracing

from the [incoherent-screaming-into-the-void] dept

Clearview AI is inserting itself into a discussion no one invited it to participate in. The discussion around contact tracing to manage and (hopefully) impede the spread of the coronavirus involves multiple governments around the world. It also includes Google and Apple, who are partnering to create a platform for contact tracing apps.

At least in the case of Google and Apple's offerings, there appears to have been a serious discussion about protecting users' privacy as much as possible while still offering a valuable service to government health agencies and people concerned about contracting the virus.

Now, Clearview has blundered into the discussion -- a company that has shown utter disdain for the millions of people it has force-fed to its multi-billion image database via social media site scraping. There is no opting in or out of this collection -- one that Clearview is selling to law enforcement agencies in the US and to government agencies around the world. If it's on the web, it's likely already in Clearview's database.

In a brief discussion with NBC News, CEO Hoan Ton-That pitched his idea for a Clearview-based contact tracing program. What Ton-That wants to do is tie his app and its database to thousands of CCTV cameras located in stores, parking lots, gyms, and other locations where, as he puts it, "there's no expectation of privacy." Ton-That doesn't explain how his system will be notified of a person's COVID status, but he's pretty sure his software will be able to recognize faces accurately. Clearview's facial recognition AI remains unproven, but it's supposedly capable of making guesses about faces using images as small as 110x110.

Ton-That also seems unconcerned about the privacy implications of adding people's health status to his enormous database of scraped personal information. He says any limitations on gathering/storage of this info would be up to whoever decides to take him up on his unsolicited offer.

Obviously, no one should do this. The AI is unproven and Clearview is far from trustworthy. Activist group Fight For The Future has issued its official statement on Clearview's contact tracing pitch. It's short but punchy.

“Absolutely the fuck not,” said Evan Greer (she/her), the groups Deputy Director.

As FFTF points out, adding Clearview to the mix just adds more potential privacy violations to the mix by dumping people's health info into a database that's already being accessed by a number of private companies and government agencies around the world. Ton-That appears to assume that anyone in a "public" area is fair game for his company. Adding thousands of cameras operated by hundreds of private companies would provide a steady flow of facial images (some tied to very personal information) into his database, making his product even more enticing to government agencies.

Hopefully, no one will let Clearview near any of this. Facial recognition adds nothing to contact tracing -- something that can be done almost anonymously using the platform developed by Google and Apple. Hoovering up facial images -- along with location data and health info -- is a privacy nightmare. And there's probably no company that should be trusted less with your personal info than Clearview, which has repeatedly demonstrated it cares nothing about the millions of people it has turned into fodder for its unproven tech.

Filed Under: artificial intelligence, contact tracing, facial recognition, hoan ton-that, privacy, trade-offs
Companies: clearview, clearview ai

Cybersecurity Firm Finds A Bunch Of Clearview's Secret Stuff Sitting Around In An Poorly-Secured Cloud Storage Bucket

from the just-scraping-it-from-the-unsecured-web-so-to-speak dept

As if we needed any further evidence that Clearview is a terrible company. The web-scraping, facial recognition provider has been pitching its unproven tech to an assortment of law enforcement agencies, one-percenters, and questionable governments for a little while now. It shows no sign of slowing down either, no matter how many people (including members of Congress) are now aware of its business practices and cheerful exploitation of billions of images found all over the web.

Someone grabbed a few internal Clearview documents and shared them with BuzzFeed earlier this year. Maybe they shouldn't have bothered. Clearview likes harvesting data and images as quickly as possible. But it's apparently less concerned with keeping its scraped stash secure from outsiders. As Zack Whittaker reports for TechCrunch, Clearview's internal files have been accessed by a security researcher, giving us yet another reason to distrust Hoan Ton-That's company.

Mossab Hussein, chief security officer at Dubai-based cybersecurity firm SpiderSilk, found the repository storing Clearview’s source code. Although the repository was protected with a password, a misconfigured setting allowed anyone to register as a new user to log in to the system storing the code.

The repository contained Clearview’s source code, which could be used to compile and run the apps from scratch. The repository also stored some of the company’s secret keys and credentials, which granted access to Clearview’s cloud storage buckets. Inside those buckets, Clearview stored copies of its finished Windows, Mac and Android apps, as well as its iOS app, which Apple recently blocked for violating its rules. The storage buckets also contained early, pre-release developer app versions that are typically only for testing, Hussein said.

If you've ever wanted to roll your own affront to humanity, Clearview helpfully left a starter kit out in the open. Of course, it's nothing without a few billion scraped images, so it's not exactly an all-in-one-kit. Maybe some Clearview insider could have hooked Hussein up with its stash of personal info. Couldn't have hurt to ask. And he could have. Included in the repository were the company's Slack tokens, which would have allowed anyone to access the company's internal communications. Also included in the storage buckets: 70,000 security cam videos of residents entering and leaving a residential building.

Hussein did disclose this issue to Clearview, but declined to take the offered bug bounty since it would have forbidden him from publicly discussing his findings. For refusing to shut up, Hussein was thanked by being called a criminal by Clearview's founder.

Ton-That accused the research firm of extortion, but emails between Clearview and SpiderSilk paint a different picture.

Lovely. Well, I'm sure this won't be the last public gaffe by the Company Most Likely To Trigger New Privacy Legislation (State or Federal). People have seen things Clearview never wanted them to see. And they've shared this stuff with the public, which now knows quite a bit about this app-based embodiment of oversharing and the damage done. It's in the midst of a very Ring-esque news cycle where every bit of new reporting makes it look even worse. But unlike Ring, it doesn't have the billions of Amazon to back it when its fortunes start to fade.

Filed Under: facial recognition, leaks, mossab hussein, security, source code
Companies: clearview, clearview ai, spidersilk

How Much Data Does Clearview Gather On People? The Answer (Sadly) Will Not Surprise You.

from the hoovering-up-your-entire-online-presence-using-this-1-simple-trick! dept

Clearview's facial recognition app links to a database of 4 billion pictures. And those photos are linked to all the data that got scraped up with them, culled (without permission) from sites like Facebook, Instagram, LinkedIn… pretty much anywhere people post photos and personal information.

There's no opting-out of this collection either, even as Clearview packages and sells access to this scraped data to law enforcement agencies in the US, as well as government agencies in countries known for their human rights abuses. Fun, fun, fun and all just a click away from exploitation by anyone with an account. That includes demo accounts operated by the super-rich and law enforcement officers told to test drive the software by running searches on friends and family members.

How much does Clearview gather on the average person? It's tough to tell. Asking Clearview directly -- at least in most of the US -- will get you nothing. However, California's privacy law (the California Consumer Privacy Act) mandates the disclosure of gathered personal data to requesters. That's what Thomas Smith of OneZero did. And here's what he got back.

The depth and variety of data that Clearview has gathered on me is staggering. My profile contains, for example, a story published about me in my alma mater’s alumni magazine from 2012, and a follow-up article published a year later.

It also includes a profile page from a Python coders’ meetup group that I had forgotten I belonged to, as well as a wide variety of posts from a personal blog my wife and I started just after getting married.

The profile contains the URL of my Facebook page, as well as the names of several people with connections to me, including my faculty advisor and a family member (I have redacted their information and images in red prior to publishing my profile here).

Based on the request process -- which involved some emailing back and forth before Clearview asked Smith to upload a photo of himself to run against its database -- Smith estimates it only took about a minute or so for Clearview to run the search and retrieve his profile.

No one knows how often Clearview crawls the web to gather new info, but what's already been gathered is impressive enough. Unfortunately, there's no one vetting this data, so it's up to end users to figure out what's right and what's wrong. The end users are (most often) law enforcement agencies, who don't really place a premium on ensuring innocent people don't get swept up in their secondhand surveillance nets.

Perhaps most worrying is the fact that some of Clearview’s data is wrong. The last hit on my profile is a link to a Facebook page for an entirely different person. If an investigator searched my face and followed that lead (perhaps suspecting that the person was actually my alias), it’s possible I could be accused of a crime that the unknown, unrelated person whose profile turned up in my report actually did commit.

Or vice versa. The negative possibilities are endless. And all of this starts with a photo of someone's face. No other input is needed. Clearview's algorithm -- which hasn't been independently tested or examined -- does the rest.

And there's more bad news. As Smith points out, Clearview's proprietary system is just it standing on looking over the shoulders of tech giants. If Clearview can do this, anyone similarly motivated can do it. One homegrown surveillance company crashing and burning from negative press just clears the way for more competitors. The bag is open and the cat is long gone. It was gone long before Clearview drifted into the public eye. A large number of companies offer social media monitoring services and plenty of other surveillance tech companies routinely scrape publicly-available data for their government agency users.

The difference is Clearview packaged its offering in an app and handed it out to whoever wanted to take a look at it. If Clearview goes down, something else will take its place. For now, it's the most visible villain. But it's not the only one. And as long as it operates, residents of California (as well as European residents under the GDPR) can demand Clearview show them how much of their data it's gathered. So, those who can should take advantage of that to let others know this is a very real threat to their privacy, not just some theoretical threat that should only cause concern for criminals.

Filed Under: data, data collection, facial recognition, privacy
Companies: clearview, clearview ai

Washington State Legislators Pass Bill Blocking Use Of Facial Recognition Tech Without A Warrant

from the over-the-sobbing-of-law-enforcement-agencies dept

We all like a good facial recognition ban, and the state of Washington is the latest to (sort of) tee one up.

The Washington state legislature passed a bill establishing new guardrails on government use of facial recognition software.

The bill cleared both chambers of the state legislature Thursday, hours before the session ended, positioning Washington as one of the first states in the nation to regulate facial recognition, a key component in the larger legal debate over artificial intelligence. The bill now awaits Gov. Jay Inslee’s signature.

Unfortunately, it's not a true ban. But it lays down a set of guidelines that will at least control its use by state agencies. The bill [PDF] requires regular reporting on use and intermittent testing for "fairness and accuracy." It also introduces something most states haven't: a warrant requirement for facial recognition use. This means facial recognition tech can't be an "always on" option for law enforcement.

The bill recognizes the threat facial recognition tech poses to the public.

Unconstrained use of facial recognition services by state and local government agencies poses broad social ramifications that should be considered and addressed. Accordingly, legislation is required to establish safeguards that will allow state and local government agencies to use facial recognition services in a manner that benefits society while prohibiting uses that threaten our democratic freedoms and put our civil liberties at risk.

It also mandates the development of data retention policies and the implementation of reporting and controls of other, non-state agency use of facial recognition tech owned by state entities. State agencies are also required to deliver the equivalent of Privacy Impact Assessments for any tech deployed.

A description of any potential impacts of the facial recognition service on civil rights and liberties, including potential impacts to privacy and potential disparate impacts on marginalized communities, and the specific steps the agency will take to mitigate the potential impacts and prevent unauthorized use of the facial recognition service.

While an outright ban (or at least a moratorium) would have been preferable, this is a positive development. It shows legislators recognize the multiple issues this tech poses and its tendency to exacerbate existing inequality problems by being seemingly incapable of accurately identifying anything other than white males.

If this bill becomes law, Washington will join a handful of cities (and one state!) around the country taking measures to protect residents from unproven tech that is swiftly becoming ubiquitous. Facial recognition tech is no longer the latest "rule of law" darling. It's increasingly being viewed as a threat to civil liberties -- not just by privacy advocates/activists, but by the people who actually can prevent their constituents from the steady creep of government surveillance.

Filed Under: 4th amendment, facial recognition, law enforcement, warrants, washington

Clearview Was A Toy For Billionaires Before It Became A Toy For Cops

from the searching-billions-of-images-for-fun-and-profit dept

Clearview's claims that its controversial facial recognition program is only for use by law enforcement agencies continues to be exposed as a lie. Documents obtained by BuzzFeed showed the company has sold its tech to a variety of private companies, including major retailers like Kohl's and Walmart.

It's also expanding its reach across the globe, pitching its products to dozens of countries, including those known mostly for their human rights violations. Even when it limits itself to law enforcement agencies, it still can't help lying -- exaggerating its success and assistance in criminal investigations.

Before Clearview became a plaything for government agencies and private corporations, it was a toy for the rich and powerful. Kashmir Hill -- who broke the first story about Clearview's problematic image-scraping operation -- has a followup in the New York Times detailing the company's unpleasant origin story.

One Tuesday night in October 2018, John Catsimatidis, the billionaire owner of the Gristedes grocery store chain, was having dinner at Cipriani, an upscale Italian restaurant in Manhattan’s SoHo neighborhood, when his daughter, Andrea, walked in. She was on a date with a man Mr. Catsimatidis didn’t recognize. After the couple sat down at another table, Mr. Catsimatidis asked a waiter to go over and take a photo.

Mr. Catsimatidis then uploaded the picture to a facial recognition app, Clearview AI, on his phone. The start-up behind the app has a database of billions of photos, scraped from sites such as Facebook, Twitter and LinkedIn. Within seconds, Mr. Catsimatidis was viewing a collection of photos of the mystery man, along with the web addresses where they appeared: His daughter’s date was a venture capitalist from San Francisco.

“I wanted to make sure he wasn’t a charlatan,” said Mr. Catsimatidis, who then texted the man’s bio to his daughter.

That's just one anecdote. There are others. Investors approached by Clearview, like venture capitalist Hal Lambert, explored the power of Clearview's app in pretty irresponsible ways. Lambert allowed his school-aged daughters access to the app. And it appears actor/investor Ashton Kutcher was given access to the app. He described an app that sounds exactly like Clearview when he appeared on the YouTube series "Hot Ones" last September.

“I have an app in my phone in my pocket right now. It’s like a beta app,” Mr. Kutcher said. “It’s a facial recognition app. I can hold it up to anybody’s face here and, like, find exactly who you are, what internet accounts you’re on, what they look like. It’s terrifying.”

It is terrifying. And far more people have had access to it than Clearview has admitted. Plenty of potential investors were given access to the app. It's not clear how many still have access, but it appears their use of the app went unmonitored/uncontrolled by Clearview. Understandably, investors want to know if the thing they're looking to invest in works, but Clearview did nothing to ensure this access was limited or used responsibly. That same attitude has carried over to its pitches to law enforcement, which encourages cops to use friends and family members as guinea pigs for tech it claims should only be used for legitimate law enforcement efforts.

Power and responsibility are supposed to go hand-in-hand. There's none of that happening here. Clearview compiled a database by scraping images from hundreds of websites and is now selling this access to pretty much anyone willing to buy it.

Filed Under: billionaires, facial recognition, john catsimatidis
Companies: clearview, clearview ai

Clearview Sued By Vermont Attorney General For Violating The State's Privacy Laws

from the engineering-its-own-downfall dept

For the second time in a little over 30 days, odious facial recognition tech supplier Clearview is being sued. Unlike the first lawsuit, which is a proposed class action over violations of Illinois' biometric privacy law, this one [PDF] is being filed by a government agency. The Attorney General of Vermont is seeking to permanently ban Clearview from collecting info about state residents or sell access to the info it's already collected.

The complaint alleges that Clearview AI has been collecting photos of Vermonters available on the internet – called “screen scraping” – and using artificial intelligence to “map” individual’s faces. Private businesses, individuals, and law enforcement use this data via an app that allows a user to identify a person within seconds using only a photograph. No Vermont state or local law enforcement agencies have used the app. Clearview AI collects the facial recognition data of Vermont children, as well as adults, without their notice or consent.

The complaint also alleges that Clearview AI violated Vermont’s new Data Broker Law by fraudulently acquiring data through its use of screen scraping.

The method used to compile Clearview's database of three billion images is going to make it very difficult for the company to dodge lawsuits like these. Scraping sites may be mostly legal, but Clearview doesn't provide any way to opt out of this collection. Scraping images from sites where users have granted the platform permission to publish their information doesn't mean the TOS agreement the user made with the platform magically migrates with the scraped data.

From the lawsuit:

[T]he term "publicly available" does not have any meaning in the manner used by Clearview, as even though a photograph is being displayed on a certain social media website, it is being displayed subject to all of the rights and agreements associated with the website, the law, and reasonable expectations. One of those expectations was not that someone would amass an enormous facial-recognition-fueled surveillance database, as the idea that this would be, permitted in the United States was, until recently, unthinkable.

Thus, when an individual uploads a photograph to Facebook for "public" viewing, they consent to a human being looking at the photograph on Facebook. They are not consenting to the mass collection of those photographs by an automated process that will then put those photographs into a facial recognition database. Such a use violates the terms under which the consumer uploaded the photograph, which the consumer reasonably expects will be enforced.

Clearview's database compilation relies on its violating website terms of service en masse. Sites have begun pushing back but at this point their legal options are limited to being angry about it. But some state-by-state litigation may actually have an effect, since it seems almost impossible for Clearview to comply with privacy laws governing the collection of biometric data. Selling access to the data makes it even more difficult, since there's no way for those packaged up in Clearview's database to request removal or limit access to the scraped data.

It's too late for Clearview to salvage its reputation by claiming it controls access carefully and sells only to law enforcement agencies to assist in solving crimes. That cat walked right out of the bag and over to BuzzFeed, which reported Clearview is selling its tech to private companies and governments engaged in ongoing human rights abuses. Claiming to serve a higher purpose while behaving like a bottom feeder doesn't do anything for anyone, least of all a company that hoped it would fly under the radar long enough to capitalize fully on a bunch of free photos it found scattered around the internet.

Filed Under: data broker law, facial recognition, privacy, privacy law, vermont
Companies: clearview, clearview ai