AI-based Street Surveillance in Hawaii (ThinkTech Tech Talks)

[music] Welcome to Think Tech Tech Talks on Think Tech Hawaii. I'm your host Jay Fidel. This episode is called AI based street surveillance in Hawaii. This is being done across the islands by companies such as Flock Security. Our guest for the show is Itila Sures of Cypac. Welcome to the show Itilla. >> Thanks for having me, Jay. I appreciate it. Always glad to help. And you [clears throat] know, I I hope that your viewers find this episode interesting because I certainly think this is an important topic. >> Let's talk about Cypac and why CYPAC is interested in AI based street surveillance in Hawaii. >> ALPR is a uh is a technology that's been around for some time and we've worked with it in different capacities. ALPR, oh man, that's a that's a mouthful, but what it really means is it's an automated license plate reader. It's super handy in service centers. So, if you're to drive your car into a service center for an oil change, they see your license plate. They're able to bring up who you are. They say, "Hey, thanks for calling, uh, coming by, Mr. Fidel, with your, you know, nice vehicle. Please, uh, pull up and, you know, we'll service your vehicle. We see that it's, you know, you have an appointment, etc. These kind of things are really useful for automatic license plate readers. The problem is that uh there's uh there are private companies that have set up inconspicuous cameras all throughout the country and they're using that same technology to gather uh everything from not just your license plate but any sort of damage you might have on your car. Uh they create heat maps of where you've been and where you're going. And uh then they also do predictive modeling to figure out where they think you might go. And uh that allows them to create a profile about you. For example, what church you go to. Uh which can then of course figure out uh what kind of religion you might ha uh might have. Uh if in on the mainland, this is a real issue with abortion uh clinics who are having uh people who are coming from out of state coming to those abortion clinics. And uh where this becomes an issue is because this is a private company, that means that that data is for sale. And our federal government uh typically would require a subpoena to do this level of surveillance on an individual or a group of individuals. And uh instead uh the loophole is that they can buy this data from a private company. And because they are buying this data from a private company that allows them to uh with some degree of confidence perform arrests. And so this has been happening for example for ICE uh to find people who are um who potentially they could deport. So an example was for example there was an individual who was of Hispanic descent. He uh went to a Home Depot to buy uh I guess uh materials to go do some sort of uh handyman work. And as he was leaving Home Depot, uh he was arrested by ICE. And it turns out that Home Depot had had some of these ALPR cameras installed in their parking lot. And uh that license plate and a description of his vehicle was sent to ICE agents who was then who were then able to find out all the whereabouts of this vehicle in the past. Then they took that information and deduced that this was an illegal immigrant and then subsequently deported him. And that's uh that's one way where you can see where private corporations uh and federal agencies are collaborating with a third party private data gathering organization to make arrests and make assumptions about people's behaviors. And by the way, it's not just Home Depot. It's Walmart and many other retailers and other kind of, you know, big box stores that are installing these uh these private ALPR devices. And of course, Flock is one of many companies that's doing this. And this is just a sign of the times. It is going to get a lot more interesting because AI is on everyone's mind. Well, guess who's doing this work? It's AI agents. And AI agents, as we know, sometimes like to hallucinate. Uh I know Jay, you're you're in you have a a law background and one of the big hot topics has been uh how AI is going to influence the legal community, right? Because it can do the research for you. However, some attorneys I've heard have become uh they got in trouble because they used AI agents as uh to gather supporting evidence and they presented it in court. But what they didn't know was that the AI agents had hallucinated facts and cases and uh when they presented it in court, it turned out to be not factual. And so AI can with a great degree of confidence say that something is true. And it sounds very believable, but it might not be. And case in point with Flock is that these ALPR readers uh were reading a uh a woman's license plate. This was a couple years ago in Colorado and she was in a parking lot with her kids and I think she was trying to find, you know, something on her phone. So she, you know, lawfully pulled over into a parking lot, was checking on her phone. Next thing she knows, she looks up. There's police all around her car. They force uh her and her kids out in gunpoint, put them on the ground, and this is all on YouTube, like you can see this for yourself. They point them all on on gunpoint and say and uh you know, put them under arrest because it turns out that her minivan happened to have the same license plate number as a stolen motorcycle from the next state away. So from few hundred miles away, someone had stolen a motorcycle and AI had somehow hallucinated that this minivan was indeed the stolen motorcycle and to go get him. And uh there's been numerous uh accounts of that where AI has hallucinated specifically under Flock and led law enforcement on these wild goose chases. And by the way, Flock is just one of many technologies out there. There are technologies that are listening devices that listen for gunshots uh which have had a very also you know controversial um success rate and it goes on and on. >> Well, that raises a lot of questions in my mind. Um you describe a a scenario where um uh uh they they know the government knows almost immediately that there's somebody in in Home Depot or wherever it is uh who might be an illegal immigrant uh almost immediately. Well, that means that it's live. It's real time. and there's somebody sitting in uh uh in in home in somebody's sitting in the government watching the screen that the screen is being tipped off uh in real time. So it's not as if they're looking at this every third day. They're looking at it right now within minutes. Am I right? >> It's even worse than that. It's not even that they're looking. is that you have AI agents looking and then based on what the AI agents think your risk level is, they'll put you on a high risk category. So, there have been examples of teenagers who have had a small uh offense, a small minor offense, and now AI agents have decided that they are a high-risisk threat and they keep sending police out to their house to harass them. So, it's not even so much like they're watching you. It's AI is watching and deciding that you are a threat based on whatever criteria it deems as uh as valid and there's NO OVERSIGHT. YOU KNOW, it sounds to me like um these cameras in Home Depot or elsewhere uh and the organizations that uh that um install these cameras are acting as quote agents end quote legally agents for homeland security. And if that's the case, if you could make that case that they're really acting for and on behalf of u of homeland security, then it's not so much a private organization, is it? Couldn't you make the argument? Are people making the argument that in fact this is the government? >> Well, that that's the loophole, right? So, the government can buy data. And by the way, this is not by far the only data gathering entity out there. Uh you and I do a lot of things on the internet, right? We post things on social media. Uh we do these kind of interviews and and we say things, right? And everything that we say is transcribed and archived and put as part of a data set about who we are. So if someone wants to know about Jidell, then it's that I promise you all the all the transcripts that you've uploaded to YouTube of all these interviews and the things that you've said, it's all been data gathered and is searchable and AI has interpreted who you are and what your preferences are and what might be important to you and what might not be important to you. maybe your health information and all that stuff is sold in real time uh for advertisers, for law enforcement is available for any for any reason. I'll give you an example. Have you ever noticed that if you've ever gone on Amazon or eBay uh perhaps you may have searched for something a week ago and then magically on your Amazon homepage there's a product that meets that demand. You know, let's say you have a backachche and you search for the you you search like why do I have a backachche? Next thing you know, two days later on your Amazon homepage, there's some, you know, backachche stuff, you know, your heating pads or Advil or whatever. How did that happen? It's because there are micro bids that occur. So, when you go to Amazon, Amazon says, "Hey, I got this guy Jay. Um, you know, what can you tell me about him?" Well, Jay has a backachche. Well, wonderful. Okay. Um, who here on the Amazon marketplace wants to uh sell Jay some Advil? Advil says, "I'd like to do that, please. I'll pay 2 cents for that." Wonderful. Okay. Now, I'll put that ad for Advil right there on uh on the homepage for Jay when he goes uh to his Amazon.com account. And that occurs in micros secondsonds in real time as you go to that site. And that's like one example that's 20 years old. This has been around for a long time. >> What about phones? Um, you know, in your notes and your comments on the in your media media appearances, you talk about phones. How do how do phones and location um location data click into this whole system? >> Well, we all love Google Maps. We all like to look at the traffic. Look at the H1 and it's green. Hey, hey, that's good, right? But then sometimes it's red. Sometimes it's yellow. Where's that coming from? Where do you think it's your phones? How fast are the phones moving on the road? And so there are some really useful um data gathering information that's real time that we get from the benefit of everyone using these phones. I love Air Tags, right? You know, if you ever travel and throw one inside one of one of those suitcases, when it ends up at the wrong airport or it's lost in luggage, hey, those Air Tags are lifesaver. It also depends on a vast uh network of millions of iPhones out there that are all talking to all the Air Tags whenever whenever uh you walk by them. So, that kind of communication can be very beneficial. Uh on my post from a couple weeks ago, uh we did talk about this and uh in my you know, in my opinion, you can't stop ALPR cameras from recording you. And if it's not going to be Flock, it's going to be some other company. Uh because they are private companies and they come up and they they you know, they're just going to keep coming. It's a startup. But one thing that you can do is change the privacy settings on your iPhone or on your Android, so that way you don't have rogue apps gathering information about you when you don't want them to. Um, have you ever noticed how sometimes you'll bring up a an app like uh maybe some game and uh you know on that uh on that Blockbuster game you you decide, oh well, why does this game need access to my contacts? Well, why does it need access to my location? Why does it need to do data sharing amongst my other apps? Like, what does it need to gather data from my pictures and my photos and my social media apps? Like, what is this game trying to do? It's harvesting your data. No such thing as a free game. It cost them money to put together. Where do you think they're going to get that back from? They're going to get it back from getting data about you and then selling it to advertisers and repeating the process. So this is an ecosystem. >> Does ALPR data integrate with the phone location data? In other words, when uh when these guys are selling data about me, um are they also selling data they got from my phone? And let me add, by the way, you should explain that this location data may be available even if my phone is off. Remember Pegasus? Remember Pegasus? Oh, there's there's been a few iterations since Pegasus for sure, but uh yeah, Pegasus, [clears throat] for anyone who's who's uh who uh who doesn't know about it, was a um was a [clears throat] package that was deployed uh by an an Israeli firm. Uh and what they were able to do was send a text message to any phone that was unpatched. It just doesn't work anymore. And essentially that one text message would allow them to completely take over the phone. So as long as they knew your phone number, they could completely take it over. And that gave them backd dooror access to be able to do whatever. And that text message of course would disappear. And they were targeting journalists and politicians and other people. And this was iPhone specific. um with Flock in particular because he asked you know where they were getting some of this data from. They can get it from data brokers. Uh so you know I mentioned earlier how a game might harvest your information and then send it off for sale. They have been purchasing data about you to combine to create that profile. And furthermore, they got in trouble because not only were they using legitimate data that was gathered and sold to them, but they were they were using illegitimate data from leaked or hacked companies. So they would find data about you on the dark web, siphon that out, and use that to build your profile. So it was it was not just purchased, but stolen data about you as well. You mentioned there are 43 plus I don't know what that translates to in actual numbers uh of these flock phones or surveillance phones in the state of Hawaii. I mean my understanding is there's there's tens of thousands or even hundreds of thousands of phones but they're real surveillance phones and they work on an internal network of you know for a given company or organization. Um this is different though this technology is different. what's the difference and how does AI play into these 43 plus phones? >> So, we're talking about the ALPR cameras that are kind of spread throughout the state of Wahoo. I was looking at the map, too. Um, and it's it's fascinating because they're not evenly distributed the way you would think. It's not like they just took these ALPR cameras and just, you know, took a took their hand and just, you know, threw it out onto a map and distributed them evenly. They're in kind of low-income neighborhoods. So, for example, in the city of Honolulu, when I checked last, there were like three of them throughout like downtown, but then I noticed in Eva Beach, there are three of them on one specific road, but PP road, which for anyone who's from Eva, it's Hal Bush, right? And Hal Bush is, as we know, is kind of a little bit of a a little rogue town. So, there's some interesting characters down there, but they put three of them down there specifically. So, what's that about? Right. So, and that's been one of the criticisms of Flock is that they're not evenly distributing these cameras to track people and maybe u you know, moderate or high-income neighborhoods. They're targeting uh you know workingclass or or or you know lower income neighborhoods and and kind of amplifying the problem because of course they are a what startup. They're what a private company. They want to show profits and and data. Well, if you're going to put together a surveillance company to surveil people, the metric of success would be either arrests or stopping crime, right? Well, if if there's a uh a a a good upper middle class neighborhood, there's probably not a lot of crime there. You want to put these cameras in the in the in the crimeridden places and then catch those criminals as as it stands, right? And those kind of activities are going to then make their product look good. So, it makes sense from their perspective why they're putting it into maybe a lower or workingclass neighborhood where there could be some of these uh things happening already. And, you know, that makes their product look good, but it it it's really u you know, biasing people. There there's also kind of like this conversation that I've um and I've read online where they talk about you know, you know, we talk about biases. Now, AI is going to inherit those biases. So, if a bunch of cameras are in, you know, low-income neighborhoods and they see all this crime, well, guess what it's going to assume? Anytime you put that camera not in a low-income neighborhood, anytime someone who matches a visual description of someone who's there, the AI engine is automatically going to assume, oh, well, that's that's a criminal, right? So, is that the kind of world you want to live in? It's it's but this is the direction we're going. The whole thing is scary. But let me let me just try to get my head straight on how the AI works. So the AI would be on the camera and the camera they use for this kind of surveillance has AI built into it. And so it is going to look for a license plate on a car, maybe even look for a face behind the windshield. Um, and it's it's going to be using AI to identify somebody who is in the language model, right? Um and then then you say that that is purchasable, transferable to some government agency or somebody who wants to buy it and are they using AI also um in identifying this person, this license plate um and you know uh examining the data, this social social networking analysis type of data um [clears throat] to put it all together Is is AI at both ends of this transaction or just one? >> Yeah, I think it's important to remember that AI is not a black box. Think of AI as doing what a human can do. So, a human would be able to look at a picture of a person driving [clears throat] their car, look at a license plate, write down that license plate number, and then say, "Oh, where where else can I look for this?" Oh, well, there's there's another spreadsheet here I have or database I have that I can put that, you know, the license plate into and then I get a name. Oh, okay. Now, now I have a name. Now, what can I find out about that name? I don't know. Let's Google search, right? Google search the name. It's called Google Derking. Put in the name. Oh, wow. I know where this person lives. His email address. Great. Oh, look. He's got a family. Oh, let's look at public records. Oh, this person just got um married last year. Interesting. Oh, and they just purchased a house. Oh, look at this. Check his credit report. >> Is this Flock you're talking about? Flock has that. >> Flock is facilitate facilitating one step of the process, right? They're the ones that, you know, they, you know, pinball machine analogy. They're the ones that pull back the pin and release the pin. Now, it all goes inside the pinball machine. The cat's out of the bag. The the pinball's bouncing around and they're checking all these things out. And at the very end of that entire chain, they could say, "Look, we got this person." And then based on our analysis, which is AI driven, right? Um, this person's a threat. And uh, all ICE has to do is say, "Hey, you know, flock, we're going to give you a million bucks for access to your database. Show us everyone who's a high threat target." And they could say, "Oh, this person is is a high threat target." Well, this is someone that we should go uh go uh go arrest. Where is he right now? Oh, he just left Home Depot. Let's nab them. >> Well, you say a million dollars. That's very interesting because, you know, it's two cents for a transaction over here, but it's a million dollars for open access. The open access troubles me. You talked about lawyers making arguments. And I would say that if I pay a million dollars and it opens it up for me, uh, then they become arguably my agent. I in the legal sense. Um, and [clears throat] all of a sudden, um, you know, flock is my agent. It's really doing the government's work. And I could make the argument in court um that that Flock must abide by my civil rights. They can't do this as a private company. They're no longer entitled to the protection of a private company. Don't you agree? >> The rules are the problem here. So the rules state that that private organizations are allowed to sell information to federal agencies and uh you know there it's just one of those kind of things and I think that there are some you know we don't know exactly how much uh is being sold to who because that flock is kind of keeping that close to their chest. Uh, but I I promise you Flock is just one of many and there's going to be more and we're in this AI space where it keeps getting better. Uh just this week for example uh Microsoft released an entire uh writeup on how hotel Wi-Fi uh conference center Wi-Fi like you know like Blazedale conference center Wi-Fi and airport Wi-Fi many have been compromised and hacked by Russian hackers and they're doing that specifically to snoop and redirect users to harvest their Microsoft 365 credentials and that was 100% thanks to AI that gave them the option the ability to do this at scale and uh of course we heard about hugging face and and uh how you know open AI got in trouble for that but just this week Meta's AI engine broke out and did the same thing. So these these AI agents are able to hack and and deliver things in in such a way that we've never seen before and we're just at the beginning of that. AI is still uh in the corral as we would say, but it is starting to break out and once it once it goes full tilt, we won't be able to stop it. You mentioned the uh interstate issue about um identifying a license plate from one state uh that was used on a motorcycle versus another state where the license plate appeared on a car. Isn't the AI smart enough to distinguish between a a mo a motorcycle and a car? Uh was that a failure, a hallucination? Um and it sounds like this is imperfect. And if I was the victim here, that is a member of the public, uh, where I was being victimized, um, I guess I should do some discovery and find out where this was taken from and whether it might be a hallucination. And if I had a claim, listen to this, Attilla, if I had a claim, I would go against a private company. It's hard to sue the United States. It's hard to sue the government, especially with this Department of Justice. But if I [clears throat] could sue Flock because Flock had a hallucination and turned it over to the government, then I arguably I could go against Flock, why wouldn't I do that? >> You can, but Flock's not the only one. And this is going to go on. >> Of course, of course. >> I I I think where where Flock has had some success, though, is that they have uh police departments paying them to [clears throat] put up these cameras. And so it's anywhere from 2500 to $3,000 per camera, uh, which is what they pay for the police department to have access to this because, uh, let's just be honest, it really does, uh, reduce local crime. Like it it can do that. Um, but like you said, the hallucinations, it's a real concern. Uh, how it's being used, that's also a real concern. Also, while we were looking here, I just double checked some facts just to make sure we're all on the same page. And yeah, it turns out that ICE gives access they they receive access to this data that we described for free. So, they're not even paying for it. >> Oh, >> yeah. >> Why why they do that as a public service? What is why do they stay in business? How do they stay in business if they doing it for free? >> It's part of their I guess community service. But one thing for example that is happening is uh I believe in Southern California if you want to open your business then you have to install cameras inside of your business and give access to uh this is a different company it's just like flock but you have to give them access uh to your to your cameras and you have to give that access to HPD and so every business from like Chick-fil-A uh to other fast food restaurants all have cameras inside that police departments have access to in real time for any reason that they want. >> Sounds like uh the protection racket in reverse. Honestly, it does. Um but you know, you talk about other states, you talk about um you know, Russia uh and hacking from afar. Seems to me that Russia might be interested in you or me and Russia could get online and bid for our information across the pond, across the miles, and they could know everything about us, where we are, where we go, what we do. I get a real profile from a company like Flock. Uh, is that happening? Or could it happen? >> It's already happened in my opinion. And I wouldn't I wouldn't look at miles anymore. The world is kind of flat. it has been for a while. Uh mass surveillance uh in the interest of public safety is always been a slippery slope. Uh so you do want to think about, you know, where's the trade-off between your privacy and uh and and safety. Uh I I think uh we've had a real challenge with uh protecting our public because the police departments have had funding issues and it's a difficult job. Let's just be honest about that. And so if there's any sort of technology that can make them safer, uh then of course why would they not use it? Uh we do the same thing for medical technology. You know the medicine has come a long way in the past few years. AI has also helped to accelerate that. Same with manufacturing. Same with software development. I mean, our world is really changing quickly. Why wouldn't law enforcement try to jump on the bandwagon and use some new technology so that they can make us safer? The problem is with all new technology, it's got bugs. >> Draconian uses um you know, malice of forethought. If I Russia wanted your information, I don't think it's got a good purpose in mind. Uh it's got a bad purpose in mind. Uh, is the government doing anything to protect us? Is anybody doing anything to protect us? Is the ACLU doing any [clears throat] is CYPAC doing anything to protect us? Well, we definitely do our best to protect those uh in the business space and that is a that is also a really big topic because uh you know what we're talking about here is uh you know personal protections uh but organizations if they are breached then it's a real problem. So this is where you know uh building your business and operating your business in a modern way is really important. Uh it's in a safe way. It's in it's in a modern way. It's in a way that's aligned with best practices to keep your organization from having a disaster. Uh so that's that's kind of our focus here. But uh these are things and you know Jay I I really like to focus on topics where you can do something about it. And if you introduce a problem like this like flock security, I mean you and I can't go up against flock and I believe that what they are trying to do is is good for us. But the the the uh the Pandora's box unfortunately has a lot of uh you know dark side effects that it's hard to mitigate. But, you know, at the same time, like, do you uh do you throw away the the baby with the bathwater? Like, do you say that we're going to live in a less secure society uh and allow people to die or get hurt uh when it could have been prevented with a piece of technology. It's it's a hard uh it's a hard line to walk, but uh I believe that if we do our if we do our part to kind of protect our individual selves by doing basic safeguards, you know, don't don't do stupid things on your phone, on your computer. Uh if if you're going to text and drive, you know, please don't put your phone in the trunk or get a self-driving car. I mean, there there's there's other options here. Uh, so I I think a common sense is going to make us a a real safe society as well. And I'm definitely not going to uh, you know, oppose law enforcement uh, for you know, using new technology to try to make us safer. >> Well, you mentioned that Flock was one of many. You've repeated that. Um, and you've mentioned that there are 43 plus cameras. I guess that's under flock or maybe it's under all of them but um there's the question of aggregate tracking and I'd like your thoughts about aggregate tracking. So, I'm company number one, call it Flack, and company number two, call it Flack 2 or somebody else. And they collaborate, they aggregate. Uh they watch you from Home Depot to uh Lowe's Hardware and back again. And and they're getting a a a picture, a signature, a location signature of you wherever you go and what you do. Even though even though Flock only has the say 43 cameras, whatever it is. So is it is it that one agent in one in one surveillance company contacts another agent and another surveillance company says, "Hey, I want to know about Aillaa. I want to know where he's been today. Can we please aggregate our data? Can we please collaborate um to make a map of where he's been uh as he passes by all of these cameras?" >> Yeah, it's all possible. And and we don't have to talk about, you know, obscure, strange uh companies. I mean, let's talk about Ring. Who's got a Ring doorbell? Lots of people. We got a Ring doorbell. So, that Ring camera feed gets uh stored in the cloud. And what if there's a crime in front of my house? Police can subpoena that video feed and they can see it. Uh, in fact, I've heard on the mainland that uh Teslas have been towed from a crime from if if a crime takes place because police show up and they say, "Oh, all these Teslas, they all have cameras that turned on when there was a uh a crime that took place here. Let's just tow all these cars and and impound them and then we'll take the video feed off the cars because it supports the crime." Now, if you're a Tesla owner, that kind of sucks for you because you show up and your car is missing, but at the same time, your car has performed a valuable service and that it's been a essentially a a a security camera on wheels and it's picked up potentially a crime that took place. So, uh again, where where do you draw the line on on this one? It's it's a tough one. >> Well, even [clears throat] if you wanted to draw a line, could you really draw a line? uh could you make a policy that uh you know regulates this uh because you know the department of justice in fact the federal government now is not as credible as it used to be. For example, if they say uh that the flock has no d no direct contact with ICE, um uh does HPD have direct contact with ICE? is is it an open, you know, an open channel between the two of them or all these companies and they're they're just collecting everything um and they're keeping it somewhere and they say HPD says there's a 90-day purge um you know on all this data, but in fact, you know, you may or may not want to believe that. Maybe it's not a 90-day purge. It's uh it's something that you have to question. and this says, "Well, you know, if you want to keep it longer than that, you got to get written approval from the the chief of police, from HPD." Really? Uh, and he has to decide that it's not being used as evidence, uh, quote evidence if it's, you know, it's going to be purged. So, um, do you believe that? Uh, and I I suggest to you that notwithstanding statements, noble statements that we're going to we're going to throw this away, we're going to purge it, we're not going to use it in the wrong way, there's nobody monitoring that. And when they say we, you know, that we'll we'll we'll purge it. That you can't you don't know if they're really going to purge it and he's got to decide that it's not evidence. Well, maybe he won't make that decision. Um what what I'm saying is that uh you know you you got to trust government to say what they're doing and uh and then maybe these days federal and state you can't and county you can't really trust what they're saying that they will do. So it's sort of an open book I suggest to you and granted it is a theoretical benefit in stopping crime, stopping violence, making our, you know, community more of a civil society and so forth, which is all good, but it just seems to me there's something very draconian about the conglomeration of all this data and the u you know the purge uh and you called it the side door loophole uh where they're not supposed to be contracts, but there's wink and blink contracts all over town about it, how the data moves and who pays who about what and who keeps it and whether you can get it and do you have to file a Freedom of Information Act or do a subpoena to find out if you got nailed improperly and so forth. I think as you say AI is just at the beginning, but so is the regulation of this kind of technology at the beginning. So that's why I'm going to make you speaker of the house. I'm going to make you speaker of the house today. I'm going to say, you're Speaker of the House. What do you do to regulate this? What what bills do you encourage? Uh what what uh statutes do you want to see adopted in order to protect the consumer to protect the ordinary Joe? protect yourself um you know from improper use and violation of your your civil rights because we have to live in a country that abides by the rule of law and we have forgotten so much about the rule of law it has been torn from us and this is part of that you'll have to agree so as a speaker of the house I mean in a in a in a in a in a in a legislature that abides by the rules that's open to protecting the public what what regulation would you try to do? >> So I believe that prevention is the best uh prevention is the best option and I'll give you an example. Uh I talked to a property manager once. Property manager came to me says you know what is the the best eviction? I said what's the best eviction? He says not letting a tenant into the property was going to be a problem in the first place. So uh this has already begun in the government space. So that means that companies who wish to participate in federal contracts are now subject to something called CMMC the cyber security model certification. That means that a third party has to come in and say look are you guys doing all the things that you say that you're doing to protect our our you know our our country's secrets and then if so you get a little check and then now you can go participate. You can go you can go play in the uh in the amusement park and [clears throat] many of our infrastructure hacks. So I'm I'm sure you hear about them like water systems, power systems, they're all being compromised because ah they just threw in these PLC controllers or they just threw in this other piece of equipment. They left the username and password default, right? Admin password, connected it to the internet. Hey, who's going to find this? Well, guess what? It can be found. especially with AI. So, um if you want to regulate companies like Flock, then having third-party assessment, having criteria, having basic safeguards and and standards, and then, you know, putting together the funding to have auditors come in and make sure that they are doing what they say that they're doing. Like for example, you said, "How do you know that after 30 days uh the video feed is being erased?" Well, an auditor could come in once a year and say, "Okay, show me show me the logs of your show me your purge logs." And uh for example, in CMMC, if you dispose of a piece of equipment that has some private sensitive data on there, you have to show that it was properly disposed of through e-waste. it was erased prior to being uh sent off to uh recycling. Right? So that same process can be applied and this is why I'm saying this is not a hard problem to solve. But it has to be done systematically. You have to have some sort of repercussions to it. And if flock wants to continue to operate as it does without oversight, that's really that's really on us because we should have said before you can put a camera up on the road, we need to we need you to demonstrate evidence that it's a configured correctly because many of these flock cameras have been hacked by uh you know some fairly techsavvy people who just drive up next to the camera and tie right into the feed and start streaming off of it because it's available. like they didn't set it up correctly or in a secure way. So if those safeguards were in place prior to them going onto the onto the streets, uh if an auditor were to come in on a regular basis and ensure that they were accurately disposing of information that they they could audit like where they get their information from and you know do all these like safeguards that they say that they're doing then that would be a much safer product because that would set the precedent for the next flock and the next one that's about to come along and Ring and all the other Amazon and the Google's and all the other ones that are jumping in on this space because let's be honest, there's a lot of money to be made here. There's some benefit to the community. We just need to make sure we do it in the right way. >> Yeah. And and as far as the individual consumer is concerned, I I'm hearkening back to a time when the military uh was trying to protect its own data, its own radio signals during a war for example. So what they do and this is in use today for various applications where they change the frequency you know 10 times a second whatever it is right I'm sure you're familiar with that so you change a frequency and nobody can really follow you for more than you know one iteration of those changes and so the same thing here suppose I came to you and I said look I got an idea my idea is I I'm going to change something in my phone I'm going to change something in my I don't know in the transmission of the data that that these guys are taking. Um and I for a dollar and a half I will protect you from that. I will give you protection against those who are sucking [clears throat] up your data and selling it. That would be that would be very interesting and attractive, wouldn't it? Uh is that doable? Is there anybody doing it? There are some uh services out there that allow you to erase or or they do their best to try to erase your digital footprint. It's called digital exhaust. So your digital exhaust is everything you do and it's kind of, you know, the that as as the name implies, it's it's kind of like a trail of what you've done. And so there are a few brokers that can, you know, help clean that up for you. Uh but again, if we had better standards in place, I mean, let's just kind of go back in history. You don't have to go very far. FDA. Where did that come from? These people were eating, you know, a a can of beans and dying the next day. Why? Because, you know, there was no standards to prevent poisonous foods from ending up on our shelves. So, the FDA, oh no, it's terrible organization. But in some ways, you know, I I can safely go to the market now, buy anything, and not be afraid of my life, right? >> Stay away from the lettuce. >> Yeah. You know, it's probably a good time to buy Taco Bell stock is all I could [laughter] say. Okay, we're really at the end of our time and I wanted to give you an opportunity to speak on behalf of Cypac. Why are you interested in this? Uh why are you, you know, communicating with the media? Why are you coming on and talking about this in in the television media and here on Think Tech? What interests you about it? And why should people call you up on the phone and ask you about it? Ask you for protection against these things that are supposed to protect us. uh because it's a really interesting double standard, if you will, a double concentration of capital standard uh that you you you want protection, but you want protection against those who are claiming to want protection. I'm getting confused already. So, what is your advice? What are your parting words to our to our audience? >> Look, it's really simple. At CIPAC, we are community defenders. Everyone who's here does the best that they can to keep our community safe. Now, most of our community is a business community, and if you're an individual, guess what? You probably work somewhere, and that business needs protection. So, our mission is to keep businesses safe here in uh in Hawaii because uh that's what keeps our island alive. Uh we're under constant threat from cyber attacks. Uh we are the gateway uh to Asia and a lot of stuff comes out of Asia. you know, we do have to do our best to prevent a disaster from occurring. We have a very high uh, you know, highly visible military presence here. So, we also have a target on our back. And, uh, you do need someone with some experience in helping you protect your business from having that kind of thing happen to you. And so, that's why we've been doing this for many years now, almost 20 years here. So, uh, you know, we we just appreciate any opportunity we can to help spread the word about what's going on in the world of technology and, uh, if your business can use some protection, this is why we're here. And of course, uh, all kinds of things happen in the news that are not related to your business protection, such as these ALPR cameras. Uh but we want to help uh you know be a source for awareness uh that we are the we are the defenders and we're we're hopefully the white knight that you turn to whenever there is an issue and if we can help we'll do our best and uh if you're definitely a business we'll do our best to prevent it from happening to you. >> That's great. Thank you for telling us about that. It sounds to me like this surveillance business falls really in the same bucket that you've been talking about, the hacking business um for a long time and it's like it's like an invasion of our private space the way all those other things are an invasion of our private space. So good for you for following it. Well, let's leave it there. Thank you so much to our guest Atilliso Cypac. I'm Jay Fidel, host of ThinkTech Tech Talks. The title of this episode has been AI based street surveillance in Hawaii and elsewhere. Thanks to our viewers for watching. We'll see you again soon for the next one. Aloha. >> Aloha. >> [music]