LPRC - CrimeScience – The Weekly Review – Episode 145 with Dr. Read Hayes & Tom Meehan

Episode Date: April 27, 2023

This week on CrimeScience, our hosts discuss some research strategies and current initiatives at the LPRC, as well as touching on AI and another set of uses and challenges for this emerging tool. Listen in to stay updated on hot topics in the industry and more! The post CrimeScience – The Weekly Review – Episode 145 with Dr. Read Hayes & Tom Meehan appeared first on Loss Prevention Research Council.

Hi, everyone, and welcome to Crime Science. In this podcast, we explore the science of crime and the practical application of this science for loss prevention and asset protection practitioners as well as other professionals. Welcome, everybody, to another episode of Crime Science, the podcast. This is the latest in our weekly update series. I'm joined by Tom Meehan, our producer, Diego Rodriguez, associate producer, Wilson Gavirino. And this particular week, Tony D'Onofrio is taking a break. He is overseas, but will rejoin us very shortly. Safe travels to Tony.

And I'm just going to take a couple minutes here to talk a little bit about some of the research that we're doing. And again, revisit a couple strategies here to help our listeners better get a grasp on what we're thinking and how we're operating at LPRC. And some of this we've talked a little bit about. But in this case, we're talking a little bit about, I talked about before, the concept of mass or concentration of power or resources. And that, you know, when it comes to the U.S. military, mass has long been the idea that who's got the most has a better chance of winning, you know, more soldiers and in later years, more aircraft or ships or bullets or whatever it might be to win the battle and eventually maybe the war, if you will. So in our case, though, mass or concentration can mean a whole lot in that we're trying

to defend ourselves, our location is, whether, again, it's a store or distribution or fulfillment center or it's an office even or a parking area. But we're trying to defend that or protect our people, defend and protect our place from theft, fraud, and violence. And so what we're doing at LPRC is taking these concepts to heart. It is really the heart and soul of what we do, and that is how do we improve the safety, the stability of an individual place so that the people that work, that shop, that deliver, or are there for legitimate reasons are truly safer and less likely to be victimized or harmed by crime by an offender or offender crew.

So that's the heart and soul of what we do. So what we've been doing is working on collaborative protection to gain some concentration. protection to gain some concentration. Because again, whether a retailer has 10 locations or 10,000 stores, for example, they are very widely dispersed and are not massed together and not going to necessarily provide mutual protection for each other. And if you look again, I can remember when I was an army officer or earlier even when I was going through U.S. Army infantry training, basic training in AIT at Fort Benning, we developed or we built fighting positions. They weren't just foxholes, but they were fighting positions. You normally had your battle buddy and there are two people in a fighting position,

but the way they were constructed is you didn't fire straight ahead. You fired just at a, you know, at a rough, let's say a roughly a 45 degree angle because the fighting positions on either side of you were doing the same. And so you create this interlocking fire, but, but direct fire couldn't hit you. Your fighting position protected you and, and really made it very unpleasant for the enemy coming at you regardless. So the idea here is that this interlocking fires, but this is this mutual protection, this collaborative, this collective protection that we're providing. And so, you know, that's what we're trying to do with the Safer Places Lab. We've got four square blocks. They represent, again, either co-located or could represent disparate locations.

How do those places work together? Let's say we're in a four square block area. We've got four different retail companies there or commercial environments. You might have some other type of, you know, have a school or a gas station or something co-located by you. How do we work together there to up the protective status, you know, profile of a given location? So again, the people now at that location and maybe nearby locations and the pathways that are used around and between are a little safer, a little more secure, better protected so we can boost what we've got. And so the concepts of mass or concentration by what we're

trying to do here is create, I wouldn't say artificial, but sustainable concentration of defenses, in this case, protective ability, and indeed a safer place, right, or places. So that's kind of the main thing. So what are some of the ways we do it? Well, can we leverage, again, what we used to call intelligence overmatch? Can we overmatch or, in other words, know more about potential offenders coming our way, more about the ways that the offenders take advantage of us? How do they or how are they currently defrauding or stealing or committing armed robberies or worse? What are the tactics they're using? What is the method of operation, the MO, right, in our case?

So the better intel we've got about who our adversaries are, who these offenders are, where they're likely to maybe go, when they might actually go there, come here to our place, how are they operating. Now we've got that and we can share that with each other. We can leverage that for the next part. And that's more precision action. Now we're not protecting everywhere from everything the same way. We might have a little more customization, might enable that store manager, that place to do certain things. If you're in a mall, a strip center, or you're co-located at intersections or a cluster of intersections nearby each other, how can you leverage the strengths and take advantage of that and cover up maybe some of your weaknesses? This company might be able to afford this mobile protective

platform in the parking lot. This other one may not. Or maybe you can share some expense. Or maybe the center owner operator, if relevant, could do the same. So what can we do with increasing the information, the intelligence, the knowledge that we need for more precision action? And how do we leverage each other's strengths? How do we work together? Maybe it's MOUs, maybe it's other ways to do it. So these are some of the things we're working on LPRC is in boosting individual place or space protection, reducing the vulnerability of that place. So we're doing that by going out beyond

what we call out in zone five, getting a better feel for what might be coming our way, who, what, when, where, why, and how, and then taking that information and getting better ourselves at how we set up our parking area, our entry-exit points, our lighting levels, or specific lighting sounds that we might employ. Again, maybe surveillance technologies and signage and combinations of all these things. Maybe it's human people, off-duty law enforcement and or security officers, third-party officers that we're leveraging, our own protective people, you know, just getting better. And then the next part is, I think, obviously leveraging the

technologies, detection technologies, like we've talked about the sensors to find out maybe potential threats online, to link together with each other, with our other stores within our chain, for example, with co-located retailers, those locations or other commercial or other places that we need to with our law enforcement partners. What are ways to leverage communication, connection, and intelligence information networks, even if it's very informal? Might be things like we hear Slack channels and Teams and just a whole host of platforms, Telegram and so on, that we might take advantage of to know more and know more have more situational awareness and even understanding With each other calls for help sharing information ideas and things like that

so those networks and connections and the tech that enables that in addition to Deterrent tech like we've talked about the from signage and and so on to place layout design And finally, it's that command and control communications intelligence that we used to call C3I, and now it's all kind of other letters. But how do we connect and control with our main decision makers, our regional decision makers, and district and market decision makers, and so on? So just wanted to kind of throw that out. These are the things that we're working on at the LPRC.

This is the way we look at things. We are, again, influencing individual offender choices, trying to convince them not to initiate against us. If they do, not to progress. Or if they continue to progress, make it less successful, fruitful. We're warning for them. And of course, document through every way we can digitally, orally, and visually so that our law enforcement and our prosecutor partners can try and take these individuals and crews out of circulation so they can't continue to hurt us. So that's it from the LPRC.

Let me go ahead and turn it over to Tom. And Tom, if you could, take it away. Well, hello, everybody. We're going to kind of do a brief recap of what's going on. I wanted to start with AI news. I think this is going to be kind of a reoccurring topic that we're going to continue to just talk about briefly on the podcast. I think it will, in fact, affect all of us everywhere in the future.

And certainly right now there's a buzz around chat GPT, but I thought it would be interesting to kind of talk about some of the things that have occurred with chat GPT and open AI. And we can use BARD. I think what I want to do here is this isn't about chat GPT for chat GPT sake. This is more about generative AI, which is the artificial intelligence, which allows things to be created, you know, to oversimplify it. Both BARD, which is Google's version, chat GPT, which is open AI or Anthropod, which would be clawed. Those are the three kind of big ones that are out right now. There are several others. All use the same type of principles for large language models. So chat GTPT, it's somewhere around 570 gigabytes or upward of 400 billion words in there. And some of the things that have occurred in the past few months with it is

a couple different instances where it gave information that was incorrect and was defamatory towards professors and other folks about things that were occurring. The model itself is not invincible to know, not invincible to errors. And I think if you go in, all of the services that are out there are showing that, you know, in very clean English, this model could be incorrect, you know, validate your information. But let's just talk about from a privacy standpoint, some of the risks or concerns that corporations have is that you're pumping your information into this device, into these tools, and you don't necessarily have a lot of control of what is occurring. So I think that's a really, really important kind of way to think about this. If you think about it that way, you're putting something in and you might not understand what really is occurring. Companies like JPMorgan Chase, Amazon, Verizon, Accenture

are banning staff from using ChatGPT. And there are more on that list. These are just kind of some of the bigger ones because of some of the privacy risks for IP as well as misinformation. Also, there are some companies that are also having growing concerns over IP, personal information, corporate information, you should limit using it. I currently use it as a more advanced search engine to answer questions quickly as someone that reads a lot. But I think there's definitely a lot to be learned. I don't think this tool is a tool that we should just write off.

I think that as in all technology and tools, there is a concern of misuse or misappropriation. Some of the cybersecurity risks with tools like this are a little bit different and aren't necessarily inherent or apparent to all of us. So with AI, one of the challenges from a risk standpoint is something called data poisoning. And that's a term that I don't think a lot of people have heard yet. But data poisoning means that AI systems rely on large amounts of data to learn and make decisions. If attackers can manipulate that data used by the systems, they can cause them to make incorrect decisions and provide false results. Data poisoning is one of the things that Elon Musk and some of the other very, very well-known people in the tech space are talking about is the ability to provide massive amounts of misinformation and basically change a narrative that exists here. misinformation and basically change a narrative exists here.

Adversarial attacks, you know, adversarial attacks are a type of cyber attack that targets AI systems. So if you were attacking a system, tricking the system into making incorrect decisions, such as misclassifying images or changing things, kind of they go similar together. And then the other one, which is I think what we always see in this space is model stealing so like someone steals that app that model and then they can use it to create their own a model one of the challenges with ai is as we

get more sophisticated at preventing attacks excuse me ai in some cases will actually enhance the ability for bad guys to use attacks. So this is not a dire warning to stop using it. It's just understand what you're doing. I think it's great to ask it questions. I think it's also really important to validate. I think it's also important to understand the differences between using Bing, using Bing as chat feature versus chat GPT versus BARD,

what they're looking at, how the backend works. Or if you're using a tool called Poe, which is owned by Query, that they have a combination of all of these. I think, as always, our goal here on the podcast is to educate. This is not opinionated. These are all factual things we're talking about. So just let's stay engaged and we'll continue to give updates here. On the global front, we talked a little bit about

geopolitical risk on the podcast. And I think when we talk about intelligence gathering, there was a couple of things that occurred over this week that are concerning. I don't know that they'll flow over to the US, but the World Health Organization has issued a warning that in the conflict in Sudan, which is almost a civil war, that an unknown group got control of a lab in Sudan. And that lab has – there's a significant risk associated with that because that lab has samples of smallpox, samples of a whole host of really, really terrible diseases that if they were accidentally released would be a challenge and if they were weaponized would be a challenge.

So there's no reason to think that they're going to be weaponized. As a matter of fact, it's probably the bigger challenge here is that if there was an accidental leak or the potential to sell this onto a third party, weaponizing viruses is not something that you just do. There's quite a bit along that goes with it. But when you think of measles, smallpox, cholera, and a whole host of really, really bad diseases in an area where an unknown Sudan fighting fashion, it could be a militia, we're not even really sure is in it, then I think that there's risk. And I do think that there'll continue to be challenges there. It is not technically a full-out civil war in Sudan, but it could be.

And then the question is in that region, which is relatively unstable and the countries around it, what happens? Why am I talking about it here? Because I believe that some of the things we learned through COVID is the supply chain is much more fragile. And if you have business in other countries that rely on certain raw materials or transports to those areas, there could be an impact, something that will certainly keep everybody apprised on. And then I just want to give a kind of an update on some of the things around TikTok. I know I talk about it all the time. About half of the folks that were polled said that they wanted to ban TikTok. One of the things about ban TikTok is it's a Chinese

owned social media network. So there's concerns of what the Chinese government could do with that information. I think there's a really interesting point here that if you went to the Android or Apple store today, you would see that there is a really, really large amount of Chinese-backed apps on the app store for both Android and Apple, we're affiliated with Chinese-owned companies. And one, not one, I'm sorry, four out of 10 of the top apps were actually affiliated with someone from the actual Chinese government being on their board. So if you think about that, that really kind of resonates with, so when you go back,

then you think about the top apps that are downloaded, four of 10 of them actually have Chinese board members. Now I'm just simply just delivering that news. I actually do quite a bit of business in China in my past, not so much anymore. And I don't think this is about China I actually do quite a bit of business in China in my past, not so much anymore. And I don't think this is about China or Chinese companies necessarily being bad. It's about the laws in China allowing the Chinese government to have untethered access to the data.

That's often what comes up with TikTok. You will find that with these other apps, that's the same case. And one of the apps is actually a cam scanner app which is a scanner app that scans documents so the risk there is like if you are scanning a document that is confidential or does have ip where does it go what what are some of the things there just a quick tip if you're using either an android or an iphone you don't need a scanning app they both have options built in so an ios an apple, you don't need a scanning app. They both have options built in. So an iOS, an Apple device, you can use your Notes app and do a scan free, embedded in there. And then on an Android application, you could actually use a Google Drive app, which is a Google app to scan a document.

So if you are using that app today or an app that you are concerned about, most of the time you're going to find that there are apps available that you don't need to necessarily use some of those apps. I often tell folks that in the real world, we're not going to have time to read the terms of services. to read the terms of services, you can usually do a pretty quick Google search or search online of your favorite search engine and get a quick, quick information of what it is. Another app that has come up quite a bit is an app called TurboVPN, which is also a Chinese-backed app where the Chinese government is on their board.

And again, the risk there is not that the app maker itself has malicious intent. It's that they are, by design, by law, having to be able to give that information. So they have to. And in recent months, there have been several other apps. A couple of apps I'm just going to throw out there are the CamScan app. You have TurboVPN and Lemon8. They're all top 10 apps. And actually, Lemon8, I think, is actually owned by ByteDance, the owner of TikTok. So this is, again, not to say you shouldn't use these apps, just thinking through what the risks are for you

and understanding that when you're downloading an app that is free, that there's got to be some sort of methodology and some sort of data for the company to monetize. Nothing is free, and the reality is here to just be vigilant about it. I think the other thing I would want to say, just switching gears a little bit to civil unrest in the US, we've had a relatively calm last couple of weeks with a couple pockets here and there, but the chatter continues to be higher than it has in the past and i think one of the things that's interesting about the open source gathering

is that what we're starting to see is again there are a lot of people that are frustrated there is division and and folks talking about it angrily doesn't always lead um torest, we'll continue to monitor that here at the LPRC and communicate it out as appropriate. And then I think the last, but certainly not least, I think I touched on this before. If you haven't had a chance to see the Loss Prevention Research Council's CBS video where both Dr. Reed Hayes and Dr. Corey Lowe were interviewed. I would encourage you to watch it. It really talks to some of the challenges that are going on and how the LPRC is helping asset protection professionals address it. And I know we'll have a lot of updates in the next upcoming weeks.

I hope to see some of the folks out in the field and some of the trade shows that are coming up in the next six weeks or so. And with that, I will turn it back over to Reid. All right. Well, thanks so much for all that good information again, Tom. We always look forward to it. And some of the areas you explore and help us think about are amazing and hugely beneficial. I want to thank our producers, Diego and Wilson. I want to wish the best to Tony and his travels. And I wish each and every one of you all stay safe, stay connected. Thanks, everybody.

Thanks for listening to the Crime Science Podcast presented by the Loss Prevention Research Council. If you enjoyed today's episode, you can find more crime science episodes and valuable information at lpresearch.org. The content provided in the Crime Science Podcast is for informational purposes only and is not a substitute for legal, financial, or other advice. Views expressed by guests of the Crime Science Podcast are those of the authors and do not reflect the opinions or positions of the Loss Prevention Research Council.