Clearview AI On Track To Win US Patent For Facial Recognition Technology (politico.com) 17
An anonymous reader quotes a report from Politico: Clearview AI has gotten the green light on a federal patent for its facial recognition technology -- an award that the company says is the first to cover a so-called "search engine for faces" that crawls the internet to find matches. Clearview's software -- which scrapes public images from social media to help law enforcement match images in government databases or surveillance footage -- has long faced fire from privacy advocates who say it uses people's faces without their knowledge or consent. Civil rights groups also argue that facial recognition technology is generally error-prone, misidentifying women and minorities at higher rates than it does white men and sometimes leading to false arrests. (A recent audit of Clearview's tech by the Commerce Department's National Institute of Standards and Technology found its results to be highly accurate (PDF), and the company said it knows of no instances to date where the technology has led to a wrongful arrest.) Now, some of those critics fear that codifying Clearview's work with a patent will accelerate the growth of these technologies before legislators or regulators have fully addressed the potential dangers.
The U.S. Patent and Trademark Office sent Clearview a "notice of allowance" on Wednesday, meaning the patent will be approved once the company pays certain administrative fees. The patent covers Clearview's "methods of providing information about a person based on facial recognition," including its "automated web crawler" that scans social networking sites and the internet and its algorithms that analyze and match facial images obtained online. "There are other facial recognition patents out there -- that are methods of doing it -- but this is the first one around the use of large-scale internet data," Clearview CEO and co-founder Hoan Ton-That told POLITICO in an exclusive interview. The product uses a database of more than 10 billion photos, Ton-That said, and he has emphasized that "as a person of mixed race, having non-biased technology is important to me." Clearview argues that there is a First Amendment right to make use of public material. "All information in our datasets are all publicly available info that people voluntarily posted online -- it's not anything on your private camera roll," Ton-That said. "If it was all private data, that would be a completely different story."
Ton-That said Clearview serves government users only and that "we don't intend to ever make a consumer version of Clearview AI." Yet Clearview says in its patent application that the invention could be useful for other purposes. The company argues that "it may be desirable for an individual to know more about a person that they meet, such as through business, dating, or other relationship." Common ways of learning about new people, like asking them questions or checking out their business cards, may be unreliable because the information they choose to share could be false, the application says. "The part that they're looking to protect is exactly the part that's the most problematic," said Matt Mahmoudi, an Amnesty International researcher who is leading the group's work to ban facial recognition. "They are patenting the very part of it that's in violation of international human rights law."
Mahmoudi of Amnesty International said that language in the patent leaves the door open to a cascade of new uses in the future. "It shows a willingness to go down a slippery slope of basically being available in any context," he said.
The U.S. Patent and Trademark Office sent Clearview a "notice of allowance" on Wednesday, meaning the patent will be approved once the company pays certain administrative fees. The patent covers Clearview's "methods of providing information about a person based on facial recognition," including its "automated web crawler" that scans social networking sites and the internet and its algorithms that analyze and match facial images obtained online. "There are other facial recognition patents out there -- that are methods of doing it -- but this is the first one around the use of large-scale internet data," Clearview CEO and co-founder Hoan Ton-That told POLITICO in an exclusive interview. The product uses a database of more than 10 billion photos, Ton-That said, and he has emphasized that "as a person of mixed race, having non-biased technology is important to me." Clearview argues that there is a First Amendment right to make use of public material. "All information in our datasets are all publicly available info that people voluntarily posted online -- it's not anything on your private camera roll," Ton-That said. "If it was all private data, that would be a completely different story."
Ton-That said Clearview serves government users only and that "we don't intend to ever make a consumer version of Clearview AI." Yet Clearview says in its patent application that the invention could be useful for other purposes. The company argues that "it may be desirable for an individual to know more about a person that they meet, such as through business, dating, or other relationship." Common ways of learning about new people, like asking them questions or checking out their business cards, may be unreliable because the information they choose to share could be false, the application says. "The part that they're looking to protect is exactly the part that's the most problematic," said Matt Mahmoudi, an Amnesty International researcher who is leading the group's work to ban facial recognition. "They are patenting the very part of it that's in violation of international human rights law."
Mahmoudi of Amnesty International said that language in the patent leaves the door open to a cascade of new uses in the future. "It shows a willingness to go down a slippery slope of basically being available in any context," he said.
Corrected title: (Score:5, Insightful)
"Privacy Rapist On Track To Win US Patent For Facial Recognition Technology"
Re: (Score:3)
I wonder how they plan to exclude EU citizens from this. They haven't asked permission to use their biometric data in this way, so they must be careful not to collect it.
Clearview AI is the fault of 'moral' objectiors (Score:4, Interesting)
Clearview AI is a perfect example of what happens when people would rather avoid being morally associated with something bad rather than work to make it better.
Amazon and google had the tech and were ideally poised to be the leading sellers of face recognition to law enforcement. No, their face recognition wasn't perfect and had bias issues but their product would have been a lot less biased and raise fewer privacy issues than the clearview product and they would have been more reluctant to sell face recognition to oppressive regimes. However, moral concerns by employees made google and amazon drop out and now we all have less privacy and law enforcement has worse, more biased (thus more likely to result in false arrests, particularly of minorities) tools.
Everyone lost.
Re: (Score:2)
there was one company that literally claimed to be founded on this principle, that someone was going to do it anyway so it may as well be them because they at least care about principles. the name of that company is Palantir Technologies.
i don't think it's all that controversial to say today that Palantir either failed or didn't actually believe in that mandate in the first place. i strongly suspect the latter but hey i'm a pessimist.
consider though: if the "legit" companies didn't have anyone criticizing t
Here is the patent (Score:5, Interesting)
The patent [uspto.gov]
The core claim: "facial image data comprising at least a captured facial image of the subject; transforming the facial image data to facial recognition data; comparing by a server device the facial recognition data to reference facial recognition data associated with a plurality of stored facial images"
Not seeing how this is anything new. How *else* has facial recognition worked for the past X years? In the details, they go on to explain that the images stored on the server are matched with personal information. Hence, the goal is to provide personal information after successful recognition. Again, hardly anything new.
Further claims are almost laughably generic. Claim 2: preprocessing. Claim 3: preprocessing may include cropping, resizing, and an laundry list of other possible techniques. Claim 4: the image is captured by a "camera-enabled" device. Ad nauseum.
Seriously, this is a patent?
Not how patents work (Score:4, Insightful)
Those critics obviously don't understand how patents work. Granting a patent will keep other companies from competing in that space, especially if the patent is overly broad or is just "this thing humans have always done, only now with computers". Patents decelerate growth of new technologies by discouraging competition and innovation (no need to improve a technology when you have a government guaranteed monopoly on it).
Patents also have no effect on laws or regulation. People can (and do) get patents for things that are highly restricted or have primarily (or completely) illegal uses.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
The patent appears to be pretty broad, almost ridiculously so, and it would seem like there would be ample prior art. It describes how most facial recognition works, and has for a couple of decades. The patent office today seems to default to granting the patent and letting the investigation fall to those who want to compete or not pay license fees. (Even I have a couple of patents, so you know, anyone can get one through.)
The key thing to read is not the patent, which is generally mostly boilerplate. The key thing to read is the claims particularly the more specific ones. It is usual to write really general initial claims to give the patent examiner something to decide against in the hope that the more specific ones will make it through
Polluting the data (Score:2)
Falls under GDPR (Score:2)
To be GDPR compliant, they would need to get consent for processing the data from each indi
Re: (Score:2)
Again, little concerns of companies when the big concerns are databases of governments and panopticons, which this misleading EU thing does not address.
Photoshop notwithstanding? (Score:3)
So, can this technology accurately determine whether or not an image has been deepfaked?
Reassuring (Score:2)
[quote]
Ton-That said Clearview serves government users only and that "we don't intend to ever make a consumer version of Clearview AI." [/quote]
Kind of like "I'm from the Government and I'm here to help you."
Patents (Score:2)
I thought patents stifle innovation. From the perspective of privacy then, is this not a good thing?