Below is a transcript of episode 97 of NeedleStack. Our hosts sit down with security and intelligence expert, Brian Kime, to discuss the implications of surveillance technology and AI on intelligence, journalism, and civil liberties.
Please note that this transcript was transcribed by AI and may include errors.
AJ Nash
Hello and welcome to another episode of Needle Stack. I'm one of your hosts. I'm AJ Nash, the digital intelligence advocate over here at Authentic8. For those who don't know me, my background, 20 some odd years in the intelligence community, chasing war criminals and fighting terrorism and those kinds of things. I've been in the private sector about 10 years now, building intelligence programs, advising intelligence-driven security. As usual, joined by co-host today, Rob Vamosi. Rob, introduce yourself.
Robert Vamosi
Sure, hi, I'm a cybersecurity expert. I'm an award-winning journalist, author of two books, and now I'm focusing on intelligence and cybersecurity.
AJ Nash
Yep, exactly. Rob's the smart one for anybody who doesn't realize it. He's the smart one of the two of us. Except today, we have somebody who might be smarter than both of us. He'll certainly tell you he is if you ask. No, I'm just kidding. Brian Kime is actually our guest today. So Brian and I are old friends. Brian's been a security and intelligence expert for decades. You'll notice for anybody who's watching as opposed to just listening, Brian's got a new gray hair in the front of his head showing just how old he is today. As I can say, we're old friends, right? That just shows experience, Brian. But no, his background's fantastic. This guy's retired. Army Reserve. He's worked at Forrester. was an analyst in the community there. He's worked for Carrier. He's got a very interesting educational background as well. He's got a master's degree in information security engineering and a master's in urban policy studies. But that was built on a bachelor's in architecture. So it comes from a really unique background. Bachelor's by way from Georgia Tech, great university. Really unique background, really smart guy. And so I'm really excited to talk today with Rob, with both of you guys, Rob and Brian about
What are we talking about today, Rob? What's your idea? What are we doing to talk about?
Robert Vamosi
So, one, talk about the rise of surveillance in our society. I know Europe's had CCTV and other things for years, but the United States, we now have Flock. We have other technologies that are tracking where we go in our cities. And I'm wondering how that either helps or hinders the intelligence community. And so, I'd like to throw that over to you guys as intelligence experts on where you see the additional data that's being collected. Does that help or does that hinder?
AJ Nash
What do you think, Brian? Let's start with you, man.
Brian Kime
Yes, yes. The short answer.
Robert Vamosi
you
Brian Kime
In general, there's just so much data. There's so many sensors out there that it's hard to move around without being tracked. And that is beneficial and harmful for both sides. So if you're the counter spy, if you are trying to track an agent in your area, if you're trying to track that case officer,
You have a ton of data. There's a ton of sensors. Like you mentioned, Rob, flock cameras. I've got one just a couple of blocks away and in law enforcement, you know, can use that to track a vehicle that may not necessarily correlate to a person, but it certainly helps if you don't have to necessarily always put a physical tail on that agent or on that case officer. can, you know, be a little sneakier. You know, you don't always have to have that physical tail. If you're using something like a flock camera.
And, but if you are, on that, if you are that intelligence officer, there's so much data out there to collect, especially when we talk about cybersecurity, you have tons of fingerprinting technologies like jar and J three J four, all those over the wire fingerprinting technologies. have everything on your phone that is tracking your location and that data is being sold and then some law enforcement or intelligence agency can acquire that data and track your device everywhere it goes. And of course, by putting all these things together, you can do some pretty interesting tracking. So it really benefits both sides, both the spies and the counter spies.
AJ Nash
Well, yeah, I think you know, it's interesting one of the challenges or what actually let's start with the advantage I guess one of the nice things about all this content right all this all this technology is Retrospectively, right if something happens, I think there's an expectation Frank with this point that will solve that that mystery that crime really rapidly right and in America Unfortunately, a lot of times that's gonna be a shooting. I mean, that's just where we are right now there's a of public shootings and and mass shootings and
It's very surprising if somebody gets far, right? Because even if you didn't see it coming, if you didn't have any heads up or any warning, whenever somebody commits one of these heinous acts, there's cameras everywhere, right? And there's whether it's the, like I said, flock cameras, whether it's, you know, public tracking, there's also the ability to, you know, grab all the content from cell towers, legally speaking with, you know, with, with proper warrants, et cetera, to start sifting through and see who was in the area, right? Plus there's all the people that are there wherever you are, there are people and they have their cameras and their phones. And, so all of this content, so it's really remarkable whenever anybody is, when it was a manhunt that lasted more than 24 hours at this point, it's pretty remarkable. now I think that also opens up a real challenge to, you know, where we're to be in the future and, and maybe we've been the past man, the conspiracy theorist opinions of, you know, manufacturing that data or manipulating that data. And then if you throw in things like AI now, right, well, it's a video, an accurate video. Did somebody create the video? Did somebody manipulate it? Somebody, you know, change it to incriminate somebody who may not have been guilty. So I think we have all this content, this technology available. So forensics are really good on the back end, right? To be able to see what happens. The proactive piece is interesting because there's so much content. First of all, obviously there are laws that prevent people from just hoovering all this stuff up and watching it all live domestically. For anybody who doesn't know, the US government doesn't do that despite what anybody might tell you. There are a lot of laws in place. It very rarely happens with mass collection. If it does, there's a lot of good reasons and a lot of legal hurdles to go through.
But this would be a challenge for that as well because now there's so much more coming in, you know, the whole hiding in the noise factor. If your content, if you're pinging of, where your cell phone is or, know, whatever, you know, where your facial recognition picture showed up, whatever it's mixed in with everybody else's, you know, there's a lot more to sift through. then of course, technology helps us sift through it faster. So it's, it's this constant effort of we need more and more and more content because that gives us a better chance of, getting the answers quicker.
But then we need more and more technology to process it because you don't have people to get through it fast enough. And then you have to trust all the technologies because you don't have time to double check. It's an endless game. I think overall we are in a much better position. I was watching a show on TV. I won't name it because it doesn't matter much, but it's set in 1850s America. And every time there's a conflict there, I'm just thinking, well, why don't they just go ahead and kill them and drag them in the woods and dump them? Like there's no forensics back then. Just make them disappear and it won't be a problem anymore.
It was so much easier back then. People died and it happened all the time. You were in the wilderness and well, Bob walked away and he never came back. Now that would not be the thought process. It shouldn't be anyway, by the way, to just kill people, but it would be much more difficult now to just go, I've had enough of this guy. I'm mad at him. I'm gonna kill him and throw him in the woods somewhere. No, you're not. No, you're not. You're gonna get caught very, very quickly by all this technology. So I think it's made us better, but it's also very complicated.
What do you think, Rob?
Robert Vamosi
Well, so we're recording this episode in the middle of December 2025. And over the last weekend, there were some dramatic violent acts committed. And one of them was in Bondi Beach in Australia. And immediately after that, somebody concocted a story that the individual who stopped the shooting for being worse than it was, was a white IT male in Australia, even though the video disagreed with that. The point that I'm making is that Ben Collins pointed out that it was absorbed by some of the AIs. And for a while, this individual that was created, his name was Edward Crabtree, propagated through these AI systems as the individual, when in fact, was Ahmed El Ahmed, who was a 43-year-old immigrant to Australia.
And where I'm going with this is that we have these AI engines picking up on misinformation and propagating it out, then comes along the investigator. And the investigator may be under a time crunch. I know there are rules and there are ways it's done. But nonetheless, it's like this first blant, what they see is what they might include in a report or inform the analysis going forward because that's what they've picked up. Is that necessarily true or has that always been the case and we're just more sensitive now because of these AI agents?
AJ Nash
Brian, you want to go first? I mean, I'll jump in, but I just feel like I just spoke for 12 minutes and felt like so I don't want to dominate the show. Brian, what are your thoughts on how AI fits in?
Brian Kime
Yeah, I mean, in surveillance, think, hand in hand, I don't think you can talk about surveillance without talking about AI and or AJ's earlier point is, can you trust the video the first reports coming out of any event? Your AI can certainly help us. You know, this podcast is called Needle Stack, right? It can help us find that needle in the needle stack. But like any data analysis, if
There's garbage in, you're going to get garbage out. And so, you know, as Rob mentioned, if someone has an erroneous report or the wrong information and gets it out first, those LLMs sometimes will pull that in first and then continue spinning out the wrong information. And the same thing goes if you're looking at what happened at Brown University, or we're talking about cyber breaches.
If the data is bad, then it's going to be hard for us to use AI to get accurate insights and inaccurate answers and then use that for whether it's indications and warnings or use it on a prosecution side of the house. So it's a mess.
AJ Nash
Yeah, I think so. I think, you know, I think to answer part of what you asked, Rob, it's always been so right. So the challenges with intelligence, with, you know, law enforcement has the same challenge is a lot of times there's that time crunch. Right. And so what we're all trained on is caveat language. Right. OK, so I don't have a lot of time. I'm going to do something really quickly. You need an initial assessment, not a final assessment, an initial assessment. I only have an hour. Good chance. Whatever I give you is going to have a lot of caveats.
you know, based on limited reporting and possible probable likelihood, you know, it's a lot of maybe it's really good professional intelligence personnel rarely jump in within an hour ago. This is who did it. Absolutely. 100%. That's actually a red flag. Somebody doing that is a good chance. You should pull them aside and find them something else to do with their time. Uh, so there's going to be a lot of caveat language early on. And so I think we're still going to see that, but I do think there's a concern.
The challenges we have with AI and AI can be a wonderful tool and really, really helpful. But there's a couple of challenges. Evolutionarily, humans are designed to find the easiest solution to problems, easiest and quickest solution. That's just how we're built. That's how we've been built from forever, right? And that's just how we're designed. So AI taps right into that, right? As opposed to going out and doing a bunch of research and trying to really find the answers and the original sources. Let me just ask this platform and oh, it'll give me sources. Will I check those?
Yeah, I will. Will everybody? Probably not. And the longer you get into it, the easier it is to be sucked into that. Even if you check sources, even if you're an expert and a professional, I got 200 sources. All right. I checked the first 12. They all seem good. I'm probably not going to check the other 188. I should, but there's a chance I'm not going to because I'm on a time crunch and somebody wants an answer. And I've got some caveat language anyway, so it's good. And so we're going to see more of that happening where people rely more and more on these platforms and we don't know what's going into them. And so you've given a really good example, you know, deserves an incredible amount of credit. He is a hero. No, no, no other word for what he did on Bondi beach. He took a couple of rounds, and saved a bunch of people's lives, and could have easily run away. He was not part of the fray. was on the, he was behind the shooters. He could have left and he chose not to. And, to have that turned into a different story that now is going to be unraveled and has been thankfully publicly very quickly. So it's been overwhelming. I'm sure the LLMs now would give you the right answer, but had that gotten a little further, had that fake story gotten further out, it gets harder to reel that back in.
AJ Nash
And what I'm really worried about is the next generation. So there was a story not long ago. It just got some, some new hubbub report came out saying 80 some odd percent of all ransomware is AI driven. that that's, listen, that report's garbage. So there's no other way to play it. It was branded by MIT. It's actually from a consortium between MIT and the private sector. And when you dig into it at all, you can see that it's garbage. There's no other word for it. I hate to be so bold about it. but there's no word for it. I mean, they included examples that, you know, were want to cry and not Petia, which was from 2017. That's for anybody who doesn't know the, the timeline here. That's five years before chat GPT came out with the first public AI. So it's very unlikely those were AI driven. I don't think the cyber criminals were sitting on multi-billion dollar technology so they could do a little ransomware attacks. And then from there, the data gets, gets worse and worse. The problem is that the product went out in April. Nobody paid a lot of attention to it early on.by the time it got some rebuke from experts in October, it had been all over the internet. And while MIT quietly just pulled that paper down, it doesn't exist publicly now. You gotta go to the archive to find it. If you Google the terms, you'll see dozens and dozens of websites that have this data, this quote, this inaccurate information, vendors and academic institutions and publishers. And that's all content that then goes into AI and trains the next round of AI and the LLMs. So,
How am I knowing that it's wrong in six months going to win an argument with somebody who says it's right when they show me all of their sources, which is all circular reporting leading back to MIT's false report that no longer exists, but how am I going to convince them that they don't know the truth? And that's going to be our problem with a lot of our research and a lot of our investigative work is all this content that's coming in has to be accurate. You know, this extra observations we're talking about and all the surveillance.
if it's manipulated and it's inaccurate and it's poured in mass into these technologies that we're going to become more more dependent on, we're going to lose sight of what truth is. And it's going to be very hard to win an argument or win your freedom in court when the preponderance of the evidence, which is all fake, but is impossible to trace back, is against you. And so that's the far end of where I think we could go with this technology, unfortunately.
Brian Kime
Yeah, I think the reality is that gen AI makes everybody a mediocre intelligence analyst, a mediocre investigator, a mediocre lawyer, you know, a mediocre doctor, whatever it is. And expertise may actually become even more valuable than it was even just a few years ago. So if we're talking, you know, surveillance,
the AI may detect, know, the case of Bondi Beach, it's gonna look at that individual and maybe it's gonna inadvertently classify that person as something they are not, but that experts, you know, will look at that and correctly diagnose, pull apart that video and accurately tell you what happened and why, where the AI just helps like a random individual kind of get it like 50%.
Brian Kime
Correct, you know, and so true expertise hopefully will come out of this and be identified and everyone can just be mediocre, but the true experts will rise above and actually, you know, help decision makers in the case of intelligence actually make better threat informed decisions in the case of, you know, digital forensics or crime scene forensics you know, actually help find killers and so forth quicker where the other, you know, 99 % of people are just looking up chat GPT or copilot or Claude or whatever. you know, throwing a few random pieces of information in there and then hoping that the math tells them what happened in the real world, which, you know, it rarely can ever do accurately.
AJ Nash
Right. Well, and Rob, I mean, we're talking about this from an Intel standpoint and from a law enforcement standpoint, but what about journalism? I mean, that's your background, right? Aren't we seeing the same issue where it's creating a bunch of mediocre amateur journalists? Is it doing the same with professionals? Do they have, do they get tricked into the laziness that affects all of us with AI and start publishing things that are inaccurate? Like how's this going to impact journalism, which we're all going to depend on as a source of truth.
Robert Vamosi
Well, in a previous episode, we talked about citizen OSINT and how easy it is for everybody to become an investigator today. The same is true in journalism. You have people that claim to be journalists and don't really have the background. And yeah, they will, they will take a story and they'll run with it without asking those questions. Like this is too good to be true. And can I get a second source to confirm that that takes too long? We don't need a second. No, you do need a second source. You really need to vet the information. And certainly when I write a book, I worry about all the things that I'm saying in the book. It's like, well, somebody is going to come back and say, that's not true, because here's this that I didn't find in the end. You know, but that hasn't happened. Fortunately, there are ways. And I think Brian and you emphasize that with good training, the expertise can actually get through all of this stuff. I'm thinking of a decade ago when the Boston Marathon bomber was caught. mean, he and his brother, and his brother were identified because it was such a public event. Everybody had their cameras out. But even there were cafes and shops that also had their cameras. within Brian, tell me a couple days, it was less than a couple days, it was very quick, they were able to sift through all of that, and identify the individuals of interest.
Brian Kime (18:29.334)
Yeah, I was thinking that too. Yeah, exactly. That, um, that, terrible bombing in 2013 was a good example of using all these CCT cameras and the power of social media. At the same time, if I recall, I think they act, some of these citizen journalists actually pointed to a couple other people that were not suspects and, and in that type of, um, kind of crowdsourced investigations.
can harm real innocent people because everyone wants to be first. And I know that's gotta be a challenge for you, Rob, as a journalist there's such an emphasis on being first with a story and not as much on being correct, you know, but expert journalism will rise to the top. And it kind of makes me want to pay for more journalism because then I know. Yeah, that I know that they are following good journalistic practices. They're not just taking random surveillance video and just throwing it out onto social media without context or, you know, claiming that it is something that isn't right. You know, I want true expertise and, know, hopefully, you know, Robin and your brethren in the journalism industry will continue to do great journalism and make it worth paying for. Right.
Robert Vamosi
Well, going back to this past weekend, in Brown University, they identified somebody, had the person's photograph, name plastered everywhere. This was from the FBI. And then they retracted it. That individual is no longer in custody, and they're now looking for somebody else. Meanwhile, on the flip side of the weekend, you had Rob Reiner and his wife murdered. of all the sources in the world, People Magazine came forward and said it was most likely his son that they're looking at. And nobody touched it for about 24 hours. And then all of a sudden, the LAPD confirmed that they were looking at the sun, etc, etc, etc. So there are good journalists out there. Sometimes they're in places like People Magazine that you least expect them to be. But they do have good sources and they vet them before they go public.
AJ Nash
And I think those are two really good examples, in my opinion, I'd like your guys' opinion on this, on why it was handled differently, right? So the university case is a case that was going to affect the public, right? There's a fear of randomness. So anybody feels like they could be a victim, this was just some students in university. Rob Reiner, not many people were gonna...
Fear that the person who killed Rob Reiner was probably on the loose looking for them next right Rob Reiner is part of a different class of people in a different community It was likely going to be personal. It was a double stabbing of a husband and wife So I wonder if that had something to do that is there's is the public Pressure right the public pressure when there's a when there's a mass shooting like a school shooting or a random attack or a bombing for instance something public like that I feel like there's a lot more pressure to get Resolution fast right to calm the public to ease the public's mind the Rob Reiner tragedy and it is a tragedy
I think it affects a lot of people emotionally, but I don't imagine there were a lot of people that thought of that and immediately thought, God, somebody's on the loose killing people. We could be next. Right. So I wonder if that doesn't allow law enforcement and journalists the right amount of time, right. To process and to work through this. Whereas it seems like when it's a much more public case, that it just feels like there's such a rush to get something to prove an effort, even if it means we brought in the wrong person. Hey, look, we're doing something. Right. And then unfortunately that person's life is turned upside down.
What do you guys think is, does it matter if it's something that feels more isolated? Does that give people more gap and time to do proper procedure and quiet as opposed to something that's a public panic where it just feels like there's pressure to get some answers out faster?
Robert Vamosi
Well, I'll take the lead on that. I think journalism has caused the problem because there's this expectation of 24-7 news. And as I know, news doesn't happen in regular fits and spurts. mean, it trickles out and then sometimes there's no news. So what do you do? How do you fill that gap? How do you keep people interested? And so it doesn't always work with journalism and we've built this sort of expectation that 24 hours a day, you can get news. Well, news doesn't always happen. And this past weekend is a great example where we had three major stories break, boom, boom, boom, one right after another. That doesn't happen very often, fortunately. there were different circumstances, as we just said, about why people were misidentified and why news waited before they actually confirmed and so forth. All different expectations.
I'm going to say that journalism kind of caused this rush to judgment and the snap decisions that we're now being asked to make. But that's my take, Brian.
Brian Kime
I think sadly politics plays a role here in each of these stories and how and it's related to journalism because I Think in you know, I apologize in advance Robin this offends you but I think some journalists and you know are are too Aligned politically with certain other groups and and come at stories current events longer term source, whatever it is from a political angle and will sometimes, oftentimes, perhaps work that story based on how it's going to affect their political tribe or the other political tribe. And so Rob Reiner, you know, while it's a terrible tragedy, I don't think politically you can spin this too much. But the other two events, Brown University and Bondi Beach in Australia, you know, both politically charged topics, stateside, you know, there's the, you know, the persistent debate over second amendment and guns rights.
And then overseas in Australia, you have anti-Semitism and immigration. And these are all very hot, passionate political topics. And so how the journalists and the professional journalists and the citizen journalists go at that current event, I think is definitely colored a lot by their politics, frankly.
Robert Vamosi
I agree with you on that. We're all liberals just putting it out there. All journalists are liberals. It's just, that's a fact and can't shy away from it. But no, that's why you have newsrooms. That's why you have different people editing and various processes. But again, it goes back to that rush to judgment being first, having something new in that 24 seven news cycle. I think that is really the problem there. AJ?
AJ Nash
Yeah, I know. I agree. I've said for years that the worst thing that ever happened to journalism was when it became a big for profit business. You know, when I was a kid, which now seems forever ago, you know, had the had the, you know, we had the what the morning newspaper in the evening newspaper and you had the nightly news and that was, you know, basically it. And I mean, I remember when CNN first started, right? So but morning paper, afternoon paper, nightly news. And I don't feel like the people who ran the nightly news were really worried about their advertising dollars. I'm sure they existed. There was obviously advertising still, but
It wasn't the same, right? And it wasn't as much competition either. You had like three networks and that was about it. And it felt like the ideal was, hey, let's get the best story out. Let's get the best news out, the most accurate news out to people, not not the story that's going to catch the most attention. You know, there weren't a lot of, I mean, headlines. Hell, now everything's a teaser even if something is important. It's you know, what's going to kill you in your own house? See us at five. It's like, my God, why don't you tell me right now something's in my house that's going to kill me, you know, and that's cliche, of course. But
Back then it was like, hey, here's the, here's the top story. Here's what's going on. Here's what you know. And it was just facts and figures and, and, and eyewitness accounts and estimates and numbers. And, they had time to do it properly. And this 24 hour cycle that you mentioned and the competition for it, there's several networks that are 24 hours a day. And then it became more about entertainment and it became more about advertising dollars. And then you stop saying things the same way, right? You may still try to get to the truth, but
how do you want to present the truth, you know, as opposed to just bold and brash and not worry about people's opinions. Cause truth is truth. It was, well, what are we going to do? That's also not going to offend the audience. How are we going to attract more audience and not lose audience? And, so I think over time we've run into a real issue with that. And every once in a while you see some, some shining examples of people that just go against that and just say, Hey, this is the facts like it or not. You know, I know our particular demographic won't care for this, but this, this is how it is. This is what we've got as evidence. And we see people that, you know, change sides, which is unfortunate that there are sides in journalism.
But I think all of this still comes back to the original discussion, which is where does the surveillance state fit in? Where does the technology fit in? And I think ideally, if we can get to the truth and get people back to just saying, hey, this is what we've seen. This is all the evidence we have and we validated it. We should be able to agree on what facts are, but I feel like we're at a point now where we don't anymore. I think the term alternative facts became a popular thing seven or eight years ago.
AJ Nash
Which by the way, there's no such thing as an alternative fact for anybody who's paying attention still. I think we started down this path of changing truth to meet whatever we want it to be. And we have people now living in alternate worlds as a result. And I think that's super dangerous and it's going to affect us more with more technology and more observation and more available evidence that people are still going to try to manipulate it to just further their argument as opposed to furthering the truth, which I think most people are just going to lose track of, unfortunately.
Robert Vamosi
So I want to turn it back to this idea of intelligence and how it's all being impacted. And I'm getting a warning message on my screen here. I don't know why. I'm back. All right. I want to take it back to the intelligence community and how surveillance is impacting that in the sense that sometimes you have to be deceptive. You have to go undercover and you have to meet in person with someone. But with all the surveillance technology, with facial recognition, with your cell phone pinging towers everywhere, et cetera, et cetera, how is somebody in law enforcement supposed to do that type of human intelligence?
AJ Nash
That's a really good one.
Brian Kime
It's very tough obviously when yeah, you are being tracked everywhere. Obviously we have online forums you can meet up, but even those can be surveilled as well. Deception and counter deception is a big part of this now and. You could take a phone and and and you could tape it to like someone's Uber or something and have them drive around. and confuse the teams that are tracking these individuals and then slip out the back door. Perhaps if the means of tracking an agent or a case officer is the phone, if you send the phone out on a trip somewhere and people are following that around, then you can walk out the back door. intelligence and deception, surveillance has all changed.
You know how we we manage our personas both online and in the physical world and how we you know conceal movement or how we fake movement It's all changed and it's you know, it's incredibly complex and You know, I think for a while, you know people mock the There there was a case was it was a couple years back where this sort of amateur ish
an individual that was traveling in a foreign country, an American, and he had like masks and wigs and stuff and you know, what was trying to kind of go old school with his, you know, clandestine meetings. But if you can, if you can fool these AI enabled cameras, you know, with some pretty cheap Hollywood type makeup special effects, you know, like
You know, maybe you can have some freedom of maneuver, of movement in a non-permissive environment like a Moscow or Beijing or something like that.
AJ Nash
Yeah, I mean, I think we've seen that there's some capabilities to fool facial recognition, right? Like you said, it's complicated. I think the era of COVID helped for some folks. Masks work really well, frankly, and the ability to be in public with a mask on and not be seen as out of the ordinary certainly can be helpful because you can mask a whole lot of your features that way. And there's a lot of different types of masks that people have used, not just the standard hospital mask. So there's that and minor prosthetic work that...
that most people can do at a costume shop is probably enough to help out. I think you got to make sure you know what you're doing on the other end too. Like you said, Brian, you do all that and then you bring your phone with you. It doesn't really do you a lot of good or you bring your Fitbit with you or any number of things that can be tracked or you use a credit card or a payment system that ties back to you. It's the full gamut. You've got to be able to be a different person. So think for those who are in that line of work, in clandestine work, I think it's
probably more challenging, but certainly within the scope of what they're trained to do, right? You just, you, have to assume the environments are more hostile than they used to be because the technologies are everywhere. Whereas you might've only limited your highest level of obfuscation for the most hostile and concerning environments. Now you have to assume that's everywhere, right? Because again, technology is everywhere. So, but I think those who are familiar with trade craft, and are skilled are probably still doing well.
You're not going to be able to get away with a fedora and a big trench coat. I don't think that's going to play out very well. can't just hide your head from the cameras like you see in TVs. But the facial recognition software is far from perfect and it doesn't take a lot to modify. still a mask and a good pair of sunglasses and a decent hat and you're probably going to make it very difficult to prove that that person is who they are and the systems won't pick them up. But you also have to make sure your technologies aren't with you.
that becomes more challenging, right? And you've got to do the same thing with whoever you're meeting with, you know, if they're under surveillance. So I think, I think all of that makes it more challenging, but I suspect, as we said earlier, it's more challenging on the backend of unraveling something after something's happened to be able to go back and find evidence. think then you're in a better position to look back and go, it turns out, yeah, these people were here and this looks suspicious. Proactively. It's very difficult. These systems aren't just out warning people, Hey, I saw this guy's face. You know, here's where he is right now. There aren't people who's actively monitoring, you know, millions of cameras all the time looking for, facial matches or anything. I think it might help you piece things together at best, but on the front end, it's still gonna be really difficult to stop people from doing clandestine activities. The technology is easier to track than the faces.
Brian Kime
Unless everyone starts wearing those AI enabled glasses that have connectivity and are scanning everyone's faces. You're just walking around town and, you know, sending that back to a database and then saying, you know, this is Rob and over there is AJ and that's scary as F.
AJ Nash
That's where you go. Yes, and that's going to be happening. I don't even disagree with you. That's not even sci-fi. We get all those meta glasses out there. You know, it's again, it's a whole nother stack of data. And yeah, if they're tied to facial recognition or they're tied to my connections, right? So so if I meet up with you for lunch, the system doesn't have to guess who you are because you're one of my connections. You're one of my contacts. So my contact list is going to be access and say, hey, Brian, this is who you're meeting. And by the way, here's his resume. And here's what he's done. You know, social media really. Here's what's new in his life. So make sure to ask him about his new.
you know, car or whatever, right? All this stuff's going to just get populated into my eyes. So I'm ready to go. Well, somebody else has all the information too. There's databases. There's something collecting that. So if everybody has these AI enabled classes on, yeah, everybody's going to know where everybody is all the time. And there's, there's somewhere there's going be content available to be accessed, making it virtually impossible because even if you're hiding from somebody, you're probably not hiding from your own people. and so it's going to be very difficult, to, hide, for somebody who, is sought out.as those technologies continue to advance. And people are going to do it because they make their lives easier.
Brian Kime
Yep. Yep. Yeah, and I see if law enforcement were to adopt these types of technologies, what does that mean in the context of the Fourth Amendment? Right? If normally on the street, I don't have to identify myself to a police officer unless like they have observed me or there's a report of me committing some crime, right? But if I'm just walking down the street and officers walk in the other direction and identifies me, is that a Fourth Amendment violation?
AJ Nash
Hmm. It's a good question. Yeah. What's probable cause now?
Brian Kime
That's tricky. That's tricky. Are we going to allow law enforcement to use this now? Typically if you're you know on the in the public street, right? You know, you don't have an expectation of privacy So I think this is going to be really interesting if those if that technology becomes you know affordable for law enforcement and and if it's adopted at all like of course, there'll be challenges to it, but I really worry You know about that If that's gonna be kosher under the Fourth Amendment. And, you know, I'm sure it'll take some, you know, some fugitives off the street, you know, but is it at the cost of, you know, our civil liberties?
AJ Nash
Yeah, it's good. Yeah. Well, imagine a simple traffic stop and an officer walks up and they have AI enabled glasses on. So it's a simple traffic stop. They don't have any probable cause to search your vehicle for anything beyond the traffic stop. had a broken taillight or you're speeding or whatever. But suddenly as they walk into the car, the glasses identify this person as somebody who has a long history of drugs, you use salient, whatever it might be. Right. Does that now become probable cause? Cause you didn't have it before. And that person hasn't displayed any activity that would have qualified as probable cause in that moment. but now you know a lot more about them, guilty knowledge because your eyes have told you, does that now create a scenario where you can search their vehicle? And again, this is assuming they're not on probation where anybody can search somebody's car is on probation automatically. Yeah.
Brian Kime
Another excellent question all alternatively the passenger right because typically at a traffic stop, right only the driver is really in scope Yeah, and now if the officer bends down he sees that you know AJ's sitting shotgun next to me and you know, and he was caught doing bad things You know a while back and you know now is that is that probable cause to go arrest the passenger or search the vehicle or
AJ Nash
They're off limits usually,
Brian Kime
All kinds of murky areas here and that technology can be pretty scary. And of course what happens when it goes wrong, right? When a human makes a mistake in the wake of Bondi Beach or Brown University or whatever, it sucks. When the AI does it, mean, one, who's responsible for that? And two, what does law enforcement do? What does the citizen do?
AJ Nash
Yeah, exactly.
Brian Kime
when they are not the person that the AI told the cop that they are, right? And then hopefully some folks are thinking about, so hopefully some lawyers are thinking about guardrails around that technology if law enforcement were to ever to use it. And I would tend to hope they would be very deliberate and slow in adopting that. And hopefully those guardrails will be in place.
AJ Nash
Yep. Yeah. just comply and hope it works out.
Brian Kime
You know, there'll be some opinions from some very knowledgeable judges and prosecutors and so forth about the admissibility of some of that surveillance data.
AJ Nash
Sure, well, then it gets back into how trustworthy is the data? Is it gonna be a mistake? Is it gonna be manipulated intentionally, right? Is there gonna be, again, anytime there's a large database of content that's useful to somebody, that means there's a couple of things. It's a target to be stolen, obviously, and it's a target to be manipulated. So, what if somebody wants to go in with the intent of manipulating data to create specific outcomes? You could have a whole class of people that find themselves being discriminated against. unintentionally, the law law enforcement, for instance, doesn't have anything to do with it. They don't know, but the data has made this happen and now a class of people have been discriminated against or a specific individual is being targeted who has, you know, has done nothing wrong because somebody, a nefarious third party manipulates that data. That's all everybody's depending on. So again, we get back into more data security and reliance on all these technologies. And we're seeing some changes right now in the laws and what's, what's allowed when you stop somebody and what's probable cause now and what's you know what is profiling. the laws are currently shifting as we're going. And if this technology is introduced, we're going to learn the hard way. think things will happen. And then afterwards we'll get to the course and find out what should or shouldn't happen because we're going to have law enforcement ahead of technology, which happens or technology, which happens. mean, this is not an unfortunate, not an uncommon thing. And so we're going to find out after the fact. And it's all happening very, very rapidly. Technology is getting in the hands of good guys, bad guys, all the above, you know, whichever side you think you're on, good or bad. and being used in ways that weren't anticipated, weren't intended, weren't understood. And so we're going to see all these consequences.
Robert Vamosi
So one of the good uses of AI is pattern matching. so I think Brian raised this in another conversation that whereas you'd be looking at satellite photos comparing two locations, the human eye can detect only so much. But then AI is going to excel at determining that, the garage door was left open in this picture, but it's closed in this other picture. It's like that type of detail is possible with a good use of AI, is it not?
AJ Nash
absolutely. I think it also can help debunk mistakes on things like time of day, for instance. think imagery analysts are very good at this in general, right? To be able look at a photo and understand what time of day it's supposed to be at, where the sun is in relation to that geography and be able look at where the shadows are supposed to be and measure that out. Imagery experts do a really good job with this, but it's still challenging in mass. Whereas AI can do that kind of work quicker and be able to say, you're saying this was at a specific location, a specific time that does not add up, that is not. how the shadows would look here. That's not where the sun was in the sky. And actually it was in this location. So there'll be a lot of opportunity to use AI to debunk mistakes or intentional manipulations as well at larger scale, right? Humans do this and do this pretty well, but it takes time. And I think the machines are going to be in a better position to do some of those things. So AI is really good at sifting through large amounts of data and helping us get to better answers on things like that rapidly. It's just when you get into the more complex. questions and answers and problem solving, that it gets more difficult and more easy to manipulate and get something in the mix that's either a mistake or an intentional manipulation. But I think you're right that the change agent, the change engine of AI, being able to recognize change quickly is something that's really, really good at. You can see a very complicated picture and quickly go, no, this, this, this, and this moved. It's something that can pick up very quickly. It's good at pattern recognition.
Brian Kime
Yeah, definitely. AI will make every geospatial analyst adequate at their job. But yeah, it really takes that expert that's been studying a target for a long time and understands all those intricacies. If you're watching a nuclear weapons facility, right? You know, what does it mean if a louver on a ventilation shaft shifts? If it changes, you know, if there's five vehicles parked out front on Thursday and there's nothing on Friday, what does that mean? That's where the cultural knowledge, that's where other knowledge about nuclear engineering and all sorts of other things come into play that that AI may not really take into account, right? But AJ, I wanna go back real quick to a point you were making about a minute ago about if we do have...
Brian Kime
law enforcement wearing these AI enabled glasses and you're collecting all this data, that data becomes a source for theft. Well, there's a, a data breach that was announced the day before we are recording this of a very popular website, that caters to people's devices and, using AI to sift through that stolen data might be interesting. and then, know, to your other point about,
AJ Nash
is it? it's a hub, is it? Okay, that cuts it down quite a bit for those who wanna know. It's a hub, is it? I gotcha, okay. I'll cut it down for some folks right there. All right.
Brian Kime
To try not to shame the victims here Rob and. And and and yeah, manipulating data like we both sides. The the group that has said they have stolen this data and the victim have admitted that there is a data theft and now yeah, can that data get manipulated and.
Well, how, if that gets into the public domain, that presents all sorts of challenges. And then now there's other plenty of metadata inside that data breach that could be enriched with other surveillance data and so many other things. think about family law cases and things like that. It's, yeah, it could be.very, very damaging for a lot of folks. And for a lot of folks that weren't actually victims, but if I just, if I decide to throw AJ's data in there and claim that, you know, he was doing whatever, right, then he's, he's got to prove that he's got to prove that he's, yeah, not guilty of that.
AJ Nash
And now I gotta spend my time defending myself. Yeah. Well, and that's a big part of it is just creating a scenario that forces people to spend time and energy to defend themselves. Even if it's completely false information at best, I got to waste time and energy proving it. At worst, there are people who will never believe me no matter what, because that's just how life works, right? Or people won't even hear the truth, right? If you'll hear an accusation, they believe the accusation, they go on about their day because we get a million headlines a day.
And it turns out months and months later that that person was exonerated, but you never hear that part of the story. And so that person, there are people who believe all sorts of things that aren't true in this world because the original headline said so, and they never followed the case to see what the result was because we don't have time and energy to do that. They're not the splashy headline. No, that splashy headline of what happened is front page. The retractions page seven in a small corner someplace, because, know, it's not newsworthy, you know, as part of it. And plus people don't want to admit it.
Brian Kime
Yep. Corrections and retractions never get the same visibility. Nope.
AJ Nash
But yeah, you talk about a breach like, you know, this this website there's also been I mean the Ashley Madison breach was I don't know what it was a decade ago or whatever. Yeah, and that's that was really embarrassing for a lot of people. For those who don't know, it's a website that caters to people who want to have extramarital affairs. And some of the folks who were breached on that one, they used government email accounts. mean, I mean, it was pretty bad all around. It's bad behavior. But you really should use your government email account to do that makes it just a little bit worse somehow. Now, for what it's worth,
Robert Vamosi
I was thinking of that.
AJ Nash
It never even occurred to me then. I didn't dig into it too deeply to begin with, but it never occurred to me then. It would have been very easy to just ram in a few other people. Hey, here's somebody I think is real jerk at work. You know, let me get part of the download of the breach and just throw in this guy's data and then send it out, you know, republish it someplace. And all of a sudden this GS 14 that I never liked working with, well, now he's in Ashley Madison data and he's got to explain himself to people and it would just cause problems. And there's so many opportunities to do that again, because there's so much data, right? There's so much coming in, you know, back to that surveillance discussion that started, whether it's video, whether it's audio, whether it's metadata, whether it's website content. There's so much out there which offers up so many opportunities to manipulate it, to create scenarios, to create stories that may not exist. It's not hard to get these data dumps off the dark web. And actually they're remarkably cheap. So it's not hard to grab this content. So you can actually have the real content if you want, and then just inject what you want. You're getting… files like it's just it's spreadsheets basically eventually and just throw in what you want and send it back out to somebody to say this is the data. It's remarkable. We haven't seen more of that frankly.
AJ Nash
Fun stuff, right, Rob?
Robert Vamosi
Wow. Yeah, poisoning breach data. Okay, that's a new low.
AJ Nash
yeah, there's always somebody out there with an idea, man. I don't think I'm the first one with that idea. I'm certain of it. I'm certain I'm not the first guy to come up with that one. It's probably even happened. I didn't even realize it. If I'm being honest, I probably processed data that had some in there and didn't know it. It just seems like it's an obvious thing for somebody to do. So of course it's happened at some point. Now the flip side is that creates a scenario where you have maybe plausible deniability if you actually are guilty of things.
Brian Kime
You mean, that is the opposite of all this with all this data out there and all the ability to manipulate and change. And this gets back to what I've said all along about the loss of truth. At what point is it just, I didn't do it. Well, there's video, there's audio, there's witnesses, there's all of this evidence. None of it's real. It's all fake. It's all AI. It's all fake news. It's all a hoax. It's all whatever. And people just lose track of what's real. And then it gets to the point of just, I'm just going to trust the person I want to trust. Even if I know I shouldn't, but you're not giving me enough evidence to.
Robert Vamosi
That's right.
AJ Nash
I shouldn't because there's no such thing as evidence anymore at that point and that's a fear I have is is it you know to begin a discussion does all this surveillance is all this content does it make it better or worse because it makes us you know better at security or worse better at intelligence or worse and as we said it's both ultimately it could be just impossible if it's gonna come down what people are willing to believe and why I like Brian's opinion that expertise is gonna have more value in a world with so much of these technologies that can't be trusted
I don't think that's what we've seen as the pattern over the last decade. Expertise is being devalued. Education is now elitism. The intelligence community now can't be trusted because we're gonna trust a foreign adversary instead because that's who some people in power wanna believe instead. Expertise has been devalued for a long time. Journalism is now, instead of having a whole set of skills and responsibilities, now it's just somebody with a camera says, I'm a journalist, and that's it. Medical science is now reduced to, conspiracy theories. I haven't seen a pattern that suggests that expertise is getting more valuable in our societies as we just rely on these technologies and people are allowed to believe what they want to believe instead of what is truth because they've got enough technologies that will tell them what they want to believe is the truth and they can argue equally with you even though it's all garbage because nobody knows what's real anymore. So all your evidence and all my evidence, they all look like evidence now. So I worry that that's going to be where we end up. And if that's the case, then yeah, this is all really, really bad. for intelligence and collection and law enforcement and just truth in general, journalism, all of it.
Robert Vamosi
Well, thank you, AJ, for your summary.
Brian Kime
Hard to disagree there.
AJ Nash
Happy holidays everyone!
Please tell me Rob you got a bright note to end on. You can't end on what I just said there. It's this dystopian future. Oh, poor Brian. Go on Brian, tell everybody good news after I just said all that.
Robert Vamosi
was just going to say, does Brian have any parting thoughts that he would like to share?
Brian Kime
No, I do agree with the folks that think all this AI is going to place a premium on true expertise. we need to, as consumers, as folks in business and in government, we need to seek out those experts and promote them and encourage a maybe return to expertise or the value of expertise because, know, infamously former director of the NSA and director of the CIA, General Michael Hayden, you know, is famous for saying we kill people based on metadata. So if we are allowing AI to find metadata that warrants certain things, we better have experts who are validating, who are contextualizing and doing the things that AI cannot do. I at the end of the day, from on the intelligence side of our conversation, an AI can tell you a lot about what has already happened, but it takes really that expert intelligence analyst with all that knowledge, all those insights in the brain and the proper trade craft to advise a policymaker or a business leader about what the threat is likely to do. next and get them to make a decision that reduces the risk of that event or threat. so, true expertise will continue to be extremely valuable and maybe more rare.
Robert Vamosi
Well, thank you, Brian, for being our guest today. This has been a great conversation. Really appreciate your expertise that you bring to it. And I want to thank our audience for listening. We appreciate all of you for joining us each week here on Needlestack. You can find transcripts and more about the show at authentic8.com/needlestack. That's authentic with the number eight dot com slash Needlestack, all one word. And be sure to let us know your thoughts. There's a comment button there where you can leave comments or you can go out on social media where we're at needle stack pod and we're found on almost all the social media platforms. And lastly, subscribe wherever you're listening or watching us today. Subscribe that way you won't miss any episodes in the future. All right, AJ I'm out.
AJ Nash
Now, I appreciate it. As you said, thanks to everyone for being listeners and watchers of the show. Thanks, Brian, for coming on. And you like all of our amazing guests. I would make this such a good show because Rob and I don't know what we're doing without you guys. And so if you're if you're enjoying the show, please let us know as as as followers. If you're not enjoying the show, please let us know. We want to make it better. Actually, the Rob, no, I don't really want to hear about it. But with that in mind, I think we'll close it out for today. Again, thank you, everybody. We appreciate your time.
This has been another episode of NeedleStack.
NeedleStack co-hosts interview professional open-source researchers who discuss tips to hone your skills, improve tradecraft, and protect yourself as you search the surface, deep, and dark web. Subscribe wherever you get your podcasts, and check out NeedleStack on the web.
Tags OSINT news
