Radio Podcast: How Can & How Should Law Enforcement Use Facial Recognition, If at all


By Montana Talks with Aaron Flint & Facial Recognition Expert / Former Texas Ranger, Skylor Hearn BLOG


<< CLICK HERE TO GO TO THE AUDIO CLIP >>

Aaron: We got a great guest here in the studio with us. Skylor Hearn is the Director of Government Affairs for a company called Clearview AI. But before that, he actually spent 27 years in law enforcement.


Not only was he the Deputy Director of the Department of Public Safety down in Texas, he also had a great career as a Texas Ranger. So we got an official Texas ranger in the studio with us right now. Skylor Hearn. Great to see you [...] So tell us about Clearview AI.


Skylor: Clearview AI is relatively new, they came into existence in 2019. They're a technology company that provides facial recognition services to predominantly law enforcement and government agencies at the state, local, and federal level for investigative purposes and national security investigations.


Aaron: We've had a great opportunity to catch up with a Democrat state representative, Katie Sullivan, out of Missoula and Republican state Senator Ken Bogner out of Mile City. And they've both been looking at it through legislative, interim committees, this whole concept of facial recognition technology.


And especially here in Montana, we're a big privacy state. We love our civil liberties. We love our privacy. It's really one bipartisan issue here in Montana where Democrats and Republicans can come together to stand up for privacy. But, what would be the benefits of facial recognition technology, especially from a law enforcement standpoint Skylor? Let's start there because I know that was one of the things where Sullivan and Bogner both said, Hey, we have clear concerns about privacy here, but we also recognize that there could be some benefits here. What are the biggest benefits to using facial recognition technology from a law enforcement standpoint?


Skylor: Yes, Aaron, I think it's important to start with the information that facial recognition technology is not new to law enforcement.


This has been used by law enforcement across this country for more than a decade back in 2008, 2009 is when some of that technology first started coming into use. And so we've been using it as a law enforcement community for a long time, it didn't become so controversial until the private sector started using it for different purposes, Google and Apple and other other folks using it more for marketing and, and customer based applications.


But in the law enforcement perspective there is digital media, a phone. Everybody has a phone. Every phone has a camera. You walk down the street. Most businesses have their own security system. There's a camera. “There's digital evidence with the majority of criminal investigations today, whether it's the liquor store security camera that got robbed, or whether it's online material being posted about, sexually explicit child exploitation videos, or human trafficking ads online, everything is digital.”


And so the positive aspect with facial recognition and really where Clearview AI fills a void in the law enforcement side is victim advocacy and identifying victims of crime. There's no other way to do that. You have your driver's license database for your state and you have your mugshot database for your state for people they've arrested. But not everybody's in there. There's no government database of six year old kids, but when you're working exclusively on a child exploitation kind of crime, And there's no, there's no suspect in that video, you don't see their face, but you see this poor child victim being assaulted and violated repeatedly.


You're trying to figure out who that child is. So then you can get a clue to where that child is, so that you can rescue them and arrest those perpetrators. Facial recognition helps significantly. In that and human trafficking.


Aaron: So basically what you're doing is you're leveraging, I'm gonna refer to it as open source and, military and law enforcement veterans understand that reference.


What that means is you're not using secret classified systems to go in and spy on people to provide these tools to law enforcement. You're using publicly available information. That's already out there on the marketplace, on the internet and using it to support law enforcement to either track down bad guys or to find a kid that maybe has been kidnapped or human trafficked?


Skylor: Yes. Correct. Everything that we provide to law enforcement in our system is available to anyone with an internet connection anywhere in the world can get to the same place. We're a search engine, just like Google. We provide the URL where that photograph is located on the open public internet. So that that lead that we provide can be investigated by law enforcement.


We don't have any biographical information in our system. Again, no names, no addresses, no dates of birth. It's just a photograph. Imagery that's out there on the, on the web.


Aaron: So I know you're based in Texas. Yes. And you, you just retired from a Texas DPS. You were a Texas ranger before you just recently retired in 2020. How, how have you seen a company like Clearview, AI? What, what are the success stories in terms of what the capability has helped deliver from a law enforcement stand?


Skylor: Again, it's significant. I got to know the company while I was still with DPS. When they first started, they came to us and asked us if we were interested in trying their technology.


And I did some of the testing on it and, and we use it then and still use it today in Texas because it fills this void specifically on these child exploitation cases and human trafficking cases and other cases where you, you have no other lead, but this image. In a lot of places around the country, what we do as law enforcement is we get on the six o'clock news and we show the picture of this unknown person.


We say, citizens of Billings or citizens of Missoula. If you know this person, call our tip line and we get 50 tips and we have to investigate those 50 tips. We're doing the same thing with technology.


Aaron: So you can point to specific numbers of cases where this has been helpful in either tracking down a bad guy or finding a lost kid?


Skylor: One of our clients that works exclusively in the child exploitation realm for last year (2021) identified 103 Victim children through online exploitation cases, tripling the number of identifications they've done previous to that. They identified four kids in three business days that were being sexually abused online in online material.


Aaron: Wow. Reminds me a lot of a couple of great Montana guys. Jeremy Mahu, a Navy seal from Whitefish, Montana. And then Nick McKinley he was an air force pararescue guy, and then he went to the CIA and they both teamed up together to create this organization, to combat human trafficking. And so what these guys did is they took all the knowledge and the tools of special operations, CIA and law enforcement and they said, okay, how can we use all of these tools that we know are available? And how can we use them to help find a missing kid? Or how can we use these tools to find, and to locate these human traffickers so that we can actually support law enforcement and take down the human traffickers here.


But I also think about it from a different standpoint, we've got the challenge of missing and murdered indigenous women here. So how could we help out in those types of efforts?


We're gonna talk about some of the privacy concerns here in a second and take phone calls as well, but Skylor, when I think about it from that standpoint, that to me really puts it in perspective. What the benefits are of leveraging things like facial recognition technology.


Skylor: You touched on a little bit earlier, but, my background at DPS as a true state police entity there, we maintained all of the crime records for the state. Every county sent their arrest records in, we were the central repository. We maintained all of the electronic fingerprints. We did all the biometric capturing that's in the criminal justice system there, and under my program, I oversaw the crime lab with all of this technology.


None of that technology has unfettered use, right? Everything has regulation and policy and statutory sideboards, if you will. We made sure that the technology is there for the right application, but the abuse or the prohibitions are also there to make sure it is governed at all levels.


But specifically at the state and local level, to ensure they are using the technology appropriately for the right things. And if somebody isn't, then those people are held accountable and appropriately for that misuse.


Aaron: Obviously there's clear privacy concerns. Here in Montana, we are a very libertarian stat and on the left and the right, the issue of privacy is one of those issues where Montanans really come together. We've got some of the strongest privacy laws in the nation. Thanks to great lawmakers like former representative Daniel Zoff.


Let's jump into our phone calls here. We got Tim in billings on the phone lines.


Tim (caller): Skylor alluded to the fact that we live in this digital world now where there's marketing, there's all this open source stuff out there. What, what kind of expectation of privacy do we have once we walk outside of our door or, or even if we sit in our own home on our own. Electronic devices? Do we really have an expectation of privacy anymore in this world?


Aaron: That's a great question. Yeah, because even if your company wasn't using this to support law enforcement, to track down bad guys or to help locate missing kids. The fact is there's all sorts of bad actors out there. There's all sorts of commercial actors who are out there, who are already using this type of stuff to target us commercially or for other nefarious purposes. What is the expectation of privacy with any of this stuff?


Skylor: The best analogy I have, again, I, I spent 30 years of law enforcement working in, in what I'll describe as the physical world. Where we pretty much know the established boundaries of private versus public inside our homes, inside our personal property. Our expectation of privacy is high. The government or anybody else doesn't come in there without a legal purpose and usually a search warrant or a court order, some lawful application to do so.


Whereas when I walk outside the house and I get on the sidewalk or I drive down the highway, my privacy expectation goes down because I put myself out where the city may have a traffic camera or somebody else taking a picture of a building, may capture me as I walk by when I'm out in public.


In the virtual world. It's not that much different. We do things every day.


Aaron: As much as we may want it to be.


Skylor: As much as we may want it to be, but it’s not, right?


Aaron: And part of that is because of our own willingness to accept this technology and to download these things on our phone.


Skylor: The efficiency of technology has gotten us to where we're using.


Apps for our banking and our medical and our prescriptions and all the things that we want to be private. We put behind a password, behind a firewall and we contract with companies to keep that information secure from other people and the government. If the government wants that information, they need to get a warrant. They need to get a court order to go in there.


But then we do TikTok and Instagram and other things that we post. We set our settings as public and we post all this information because we want people to see it.


Aaron: And if I understand it correctly, I know the concern Clearview AI has is, you want to be able to provide these tools to law enforcement right now. But there are basically no guardrails in place regarding facial recognition technology.


So I think to their credit, these lawmakers, Democrat and Republican alike, want to put some guardrails in place. But your fear and maybe the fears of some in law enforcement is that instead of just putting left/right limits in place and guardrails in place, they may just put a ban on it all together. Talk to us about why you'd rather see some guardrails rather than an all out ban.


Skylor: Of course and that's why I'm here in Montana this week, trying to provide some context and facts. There's a lot of misinformation about the technology out there so I’m here to hopefully help those legislators and Montanas themselves decide where they want the side rails to be.


Facial recognition is not a confirmatory biometric, it's not DNA. It's not a fingerprint. So it should not be the sole basis for an arrest. And you put that in statute. That's one of your side boards. You make that clear. It is not probable cause by itself. You put in statute, another side board is law enforcement cannot use it just to identify someone doing a constitutionally protected activity. If there's no, there's no probable cause.


Aaron: So if you're out there protesting on the steps of the state capital and it's a gun rights rally or something else, they can't go use facial recognition to try to identify, to figure out who you are.


This was a message we got from Dennis in Bozeman. On our Montana talks app, Dennis says a warrant should be needed to utilize facial recognition for law enforcement purposes to secure our fourth amendment rights, to be secure in our personal effects. So a warrant should be needed in order to leverage facial recognition technology.


Skylor: I'll have to disagree with Dennis on that because the fourth amendment and the warrant requirement is about how you acquire information, how you access it. It's not about how you process it. And I think that's getting mixed up in some of the talk in the legislative effort about the warrant and the court order how law enforcement gets your information may require a warrant.


If I have to search your house as part of a crime, I need to get a warrant to do that, to acquire the evidence. Once I have the evidence lawfully, how I process doesn't have the fourth amendment apply to that. And it doesn't apply to that anywhere in the country. There is no precedent for a warrant requirement for a tool.


Aaron: Well and I wonder how that would work. Let's say you are going back to the missing kid scenario. You have a kid that goes missing, in Wolf point Montana. And we want to find that kid as quickly as possible. We're scouring thousands and thousands of images on the open source internet to try to see where this kid could possibly be. And so, do you have to do 15,000 warrants when you're looking for a missing kid? So that's where I could see that it's challenging depending on how you write that into the law.


Skylor: Its not applicable to this part of the discussion. There is no precedent for that in the United States. My analogy would be if I took a thousand photographs and laid them out on your desk here and I started looking through them myself manually. And said, okay, I think these three are pretty close to the person I'm gonna investigate these three. I have done it myself.


You're saying that I can do that without a warrant, but if I use a tool or technology to do that same thing for me now, I need a warrant to do it.


[...]











Note: The following is the output and edited transcribing from an audio recording. Although the transcription is largely accurate, in some cases it is modified due to transcription errors, clarifying pronouns, summarizing or otherwise making the written version clear to the reader. It is posted as an aid to understanding the radio podcast, but the full audio can be listened to on this page instead, or in addition to.

___


AARON FLINT

Montana Talks

“I host Montana's leading statewide radio talk show "Montana Talks." Aside from serving in Iraq and Afghanistan with the military, I also spent over two decades working in Montana politics and broadcasting. This included a stint as the State Director for Ryan Zinke’s Congressional offices in Montana, and as communications director for the gubernatorial campaign of current Governor Greg Gianforte. Prior to that, I grew a statewide radio show in Montana from 10 signals to more than 20 across the state. My work has been featured nationally by Fox Business, The Drudge Report, Huffington Post, Politico, and others. As an officer in the Army National Guard, I served four military tours overseas including time in Iraq, Afghanistan, the Horn of Africa, and the Arabian Peninsula. I enjoy combining my real-world background and Montana roots, to hopefully give listeners a side of the story they don't get to hear anywhere else.”



SKYLOR HEARN

Director of Government Affairs for Clearview AI & Lt. Colonel (Ret.), Texas DPS

Skylor Hearn is Director of Government Affairs for Clearview AI. Skylor has more than 27 years of progressive law enforcement experience including as former Texas Ranger Captain and DPS Deputy Director. Throughout his career, Skylor provided leadership and direction to patrol, investigations, and support operations across large geographic regions. Additionally, he led highly specialized investigative units and programs including Texas Ranger Company “G” and “D”, criminal justice data collection and analysis, the forensic crime laboratory system, biometric data systems, and law enforcement and civilian training.


In 2022, Skylor began working for Clearview AI where he utilizes his law enforcement expertise and legislative experience to help policy makers and law enforcement executives craft model use policies and laws nationally for the utilization of facial recognition technology and public online content by law enforcement.


Skylor remains commissioned as a Special Texas Ranger by DPS and has been recognized with the DPS Director’s Citation for Criminal Interdiction as well as received the FBI Director’s Award for Rescue of a Kidnapped Child.