WEBVTT 1 00:00:00.160 --> 00:00:05.719 Welcome to our deep dive today. Yeah, it's a fascinating one. 2 00:00:05.919 --> 00:00:09.439 We're diving deep into the world of cyber warfare and 3 00:00:09.560 --> 00:00:10.759 hardware trojans. 4 00:00:10.759 --> 00:00:12.560 Oh yeah, very relevant. 5 00:00:12.800 --> 00:00:15.919 And you have provided us with quite a mix of sources. 6 00:00:16.480 --> 00:00:20.399 We've got excerpts from books about viruses, hardware software, trojans, 7 00:00:20.719 --> 00:00:25.879 and some potential countermeasures, and then also some articles, yeah, 8 00:00:26.000 --> 00:00:27.120 research papers as well. 9 00:00:27.160 --> 00:00:29.239 So it seems like you're really trying to get a 10 00:00:29.280 --> 00:00:30.640 good grasp of this whole thing. 11 00:00:30.760 --> 00:00:35.320 You're looking to understand this complex topic, right, not just 12 00:00:35.359 --> 00:00:39.479 what these threats are, but how vulnerable our everyday tech 13 00:00:39.520 --> 00:00:40.679 really is exactly. 14 00:00:41.079 --> 00:00:44.240 And it's a really good question to be asking. You know, 15 00:00:44.399 --> 00:00:48.159 most people when they think of cyberwarfare, they maybe think 16 00:00:48.159 --> 00:00:50.640 of something out of a movie, right, But the reality 17 00:00:50.679 --> 00:00:53.000 is it's a lot more complex and a lot closer 18 00:00:53.039 --> 00:00:54.479 to home than we might like to think. 19 00:00:54.560 --> 00:00:56.640 Oh. Absolutely. And you know what's so interesting to me 20 00:00:57.240 --> 00:00:59.880 is how this all evolved. Yeah. Well, we think about 21 00:01:00.000 --> 00:01:03.280 cyber warfare, right, think about you know, nation states and 22 00:01:03.399 --> 00:01:08.239 shadowy government agencies. But our sources suggest a very different 23 00:01:08.239 --> 00:01:09.079 origin story. 24 00:01:09.280 --> 00:01:13.760 Yeah they do. It's it's not what you would expect. Now, 25 00:01:13.799 --> 00:01:16.760 what do you think it's really interesting, Yeah, tell me more. 26 00:01:17.159 --> 00:01:21.200 Criminal organizations were actually some of the earliest innovators in 27 00:01:21.280 --> 00:01:21.959 this space. 28 00:01:22.400 --> 00:01:26.319 So they were using software exactly and hardware trojans to 29 00:01:26.400 --> 00:01:31.200 commit crimes way before, way before government's caught on. Governments 30 00:01:31.200 --> 00:01:32.280 even knew it was going on. 31 00:01:32.879 --> 00:01:33.879 Yeah it was. 32 00:01:34.000 --> 00:01:36.400 That's so interesting. Wow, what kind of crimes are we 33 00:01:36.439 --> 00:01:36.959 talking about? 34 00:01:37.040 --> 00:01:40.840 Give you an example, early days of online banking, there 35 00:01:40.879 --> 00:01:46.040 was this group they developed a hardware trojan. It was 36 00:01:46.120 --> 00:01:50.079 disguised as a simple USB card reader. Oh wow, So 37 00:01:50.120 --> 00:01:53.719 you can imagine these unsuspecting customers. They plug it into 38 00:01:53.799 --> 00:01:57.599 their just to check their balances, you know, and bam, 39 00:01:57.640 --> 00:02:02.920 this trojan stealing their credentials, steals their login credentials. 40 00:02:03.200 --> 00:02:04.680 Oh my gosh, gives. 41 00:02:04.599 --> 00:02:09.479 The criminals access to their accounts. That's really, of course illegal, Yeah, 42 00:02:09.639 --> 00:02:11.599 but a brilliant use of this technology. 43 00:02:11.840 --> 00:02:16.000 Yeah wow. So how did this jump from stealing a 44 00:02:16.039 --> 00:02:20.879 few bucks to becoming full blown cyber weapons. 45 00:02:20.919 --> 00:02:24.960 Well, intelligence agencies and militaries, Oh, they saw the potential 46 00:02:25.159 --> 00:02:29.280 of these techniques. They quickly recognize the power of what 47 00:02:29.319 --> 00:02:30.960 we call information weapons. 48 00:02:30.960 --> 00:02:34.080 Okay, so information weapon, Yeah, how is that different from 49 00:02:34.120 --> 00:02:35.159 a traditional weapon? 50 00:02:35.360 --> 00:02:36.879 Let me try and break that down. So, think of 51 00:02:36.879 --> 00:02:41.479 a traditional weapon like a bomb causes physical damage, right, 52 00:02:41.479 --> 00:02:46.280 an information weapon it targets data, software, and hardware with 53 00:02:46.319 --> 00:02:49.080 the aim to disrupt, manipulate, or destroy. 54 00:02:49.439 --> 00:02:53.000 So the goal is to undermine Yeah, the target's capabilities, 55 00:02:53.039 --> 00:02:57.159 that's right, And that can have real world consequences very 56 00:02:57.199 --> 00:02:59.719 much so. So it's like a virtual attack, yeah, but 57 00:02:59.759 --> 00:03:01.280 with real world cost. Yes. 58 00:03:01.599 --> 00:03:05.680 And just like with traditional weapons, information weapons can be 59 00:03:05.759 --> 00:03:08.840 offensive to attack or defensive to protect. 60 00:03:09.400 --> 00:03:10.360 So give me an example. 61 00:03:10.759 --> 00:03:14.639 Okay, so what does a real world cyber attack look like. 62 00:03:14.919 --> 00:03:15.240 Yeah. 63 00:03:15.360 --> 00:03:19.319 One of the most notorious examples is the stuxnetworm. Okay, 64 00:03:19.360 --> 00:03:25.199 this is back in twenty ten. This targeted Iran's nuclear program. 65 00:03:25.280 --> 00:03:26.520 Okay, I've heard of this one. 66 00:03:26.639 --> 00:03:28.199 It is a fascinating case study. 67 00:03:28.400 --> 00:03:31.159 So this is the one that actually caused physical damage. Yes, 68 00:03:31.319 --> 00:03:33.120 I remember reading about this. This is crazy. 69 00:03:33.159 --> 00:03:37.439 So Stucksnet specifically targeted these centrifuges that are used to 70 00:03:37.560 --> 00:03:43.360 enrich uranium. It manipulated the programmable logic controllers Okay, which 71 00:03:43.400 --> 00:03:46.159 basically are the computers that control the centrifuges. 72 00:03:46.280 --> 00:03:46.599 Got it. 73 00:03:46.719 --> 00:03:49.439 The worm caused them to spin out of control. Oh wow, 74 00:03:49.520 --> 00:03:50.599 damaging the equipment. 75 00:03:50.879 --> 00:03:55.000 So it set back Uron's nuclear program significantly. 76 00:03:55.199 --> 00:03:56.439 Yeah, significantly. 77 00:03:56.479 --> 00:03:57.240 That is incredible. 78 00:03:57.280 --> 00:03:59.919 It really highlights I mean, not incredible in a good way, right, 79 00:04:00.080 --> 00:04:00.680 of course not. 80 00:04:00.840 --> 00:04:03.319 But it just shows how powerful these things are. But 81 00:04:03.360 --> 00:04:05.599 we're here to talk about hardware trojans. 82 00:04:05.280 --> 00:04:07.000 Right right, specific So. 83 00:04:06.960 --> 00:04:09.000 Why are those considered such a big deal in this 84 00:04:09.080 --> 00:04:11.199 whole cyber warfare landscape. 85 00:04:11.360 --> 00:04:14.919 Well, hardware trojans are especially concerning because of how stealthy 86 00:04:15.039 --> 00:04:19.319 they are. Okay, they are literally embedded within the hardware itself, 87 00:04:19.839 --> 00:04:22.319 making them incredibly difficult to detect. 88 00:04:22.480 --> 00:04:27.439 So like if traditional software like anti virus can't catch them, yeah, 89 00:04:27.480 --> 00:04:27.879 what can? 90 00:04:28.319 --> 00:04:28.959 A good question. 91 00:04:29.040 --> 00:04:31.160 You almost need to know exactly what you're looking for. 92 00:04:31.360 --> 00:04:32.319 Yeah, you really do. 93 00:04:32.360 --> 00:04:34.959 And even then, it's a needle in a haystack, it 94 00:04:35.040 --> 00:04:35.560 really is. 95 00:04:35.680 --> 00:04:40.199 It's finding these tiny pieces of malicious circuitry hidden inside 96 00:04:40.439 --> 00:04:43.560 legitimate hardware, right, it's a huge challenge. 97 00:04:43.639 --> 00:04:43.959 Yeah. 98 00:04:44.000 --> 00:04:47.399 They're designed to blend in Oh wow, often disguised as 99 00:04:47.439 --> 00:04:48.519 harmless components. 100 00:04:48.600 --> 00:04:50.040 So give me an example, like a. 101 00:04:49.920 --> 00:04:55.160 Phone charger, okay that secretly records your conversations. Oh my gosh, Okay, 102 00:04:55.199 --> 00:04:57.279 that's the level of threat that we're talking about here. 103 00:04:57.360 --> 00:04:59.560 Hold on back up for a second. Yeah, how does 104 00:04:59.600 --> 00:05:02.439 it you can even get into a phone charger in 105 00:05:02.480 --> 00:05:04.759 the first place. Well, that's the thing, Like, isn't that 106 00:05:04.800 --> 00:05:07.399 all made in like these really high tech factories. Yeah, 107 00:05:07.759 --> 00:05:10.759 you'd need access to the manufacturing process to do something 108 00:05:10.800 --> 00:05:11.120 like that. 109 00:05:10.959 --> 00:05:11.839 That's exactly right. 110 00:05:11.920 --> 00:05:13.000 So it's an inside job. 111 00:05:13.360 --> 00:05:17.199 You could say that hardware trojans are typically inserted during 112 00:05:17.199 --> 00:05:20.360 the manufacturing process. It could be like a tiny, almost 113 00:05:20.439 --> 00:05:24.480 invisible modification to a chip's design, a few lines of 114 00:05:24.519 --> 00:05:27.879 code added to the firmware, or even like a completely separate, 115 00:05:28.040 --> 00:05:30.279 malicious circuit added during assembly. 116 00:05:30.560 --> 00:05:33.480 That's scary. It is a concerned like almost any device 117 00:05:33.480 --> 00:05:34.279 could be compromised. 118 00:05:34.279 --> 00:05:35.759 Then, Yeah, that's the scary part. 119 00:05:35.800 --> 00:05:37.839 It's like a ticking time bomb waiting to. 120 00:05:37.839 --> 00:05:39.759 Go off, you could think of it that way. 121 00:05:39.839 --> 00:05:40.680 And they're everywhere. 122 00:05:40.920 --> 00:05:45.439 And that's what makes it even scarier, is this potential pervasiveness. Yeah, 123 00:05:45.480 --> 00:05:53.759 we're talking vulnerabilities in everything from smartphones and laptops to cars, airplanes, 124 00:05:53.920 --> 00:05:57.040 critical infrastructure. Wow, it's a big problem. 125 00:05:57.079 --> 00:05:59.920 Okay, I am officially freaked out now. Yeah, we've got 126 00:06:00.040 --> 00:06:03.600 these tiny little trojans lurking in all of our devices. 127 00:06:03.800 --> 00:06:05.279 Yeah, potentially everywhere. 128 00:06:05.399 --> 00:06:07.279 They could do all sorts of an anti things. 129 00:06:07.319 --> 00:06:08.279 It's a big problem. 130 00:06:08.439 --> 00:06:09.600 But how do they actually work? 131 00:06:09.879 --> 00:06:10.279 Yeah? 132 00:06:10.319 --> 00:06:12.279 How do you technically speaking. 133 00:06:12.199 --> 00:06:13.560 Hide something like that in a chip? 134 00:06:13.600 --> 00:06:15.639 How do you hide it? 135 00:06:15.639 --> 00:06:19.879 It's incredibly complex. Yeah, it really speaks to the ingenuity 136 00:06:20.000 --> 00:06:23.800 of the attackers. You know, these chips that we have 137 00:06:23.839 --> 00:06:28.800 today are so complex, billions of transistors interconnected in ways 138 00:06:28.800 --> 00:06:31.920 that are almost impossible to fully comprehend. 139 00:06:32.079 --> 00:06:34.959 Yeah. You've got engineers who designed these things, and even 140 00:06:35.000 --> 00:06:35.800 they even. 141 00:06:35.680 --> 00:06:39.199 They don't always understand every single detail how it works. Yeah. 142 00:06:39.240 --> 00:06:43.160 So it's like hiding a single altered pixel in a 143 00:06:43.240 --> 00:06:45.240 massive high resolution image. 144 00:06:45.319 --> 00:06:46.639 Yes, that's a great analogy. 145 00:06:46.800 --> 00:06:47.720 No one would ever notice. 146 00:06:47.800 --> 00:06:48.639 Nobody would notice. 147 00:06:48.680 --> 00:06:50.439 It's just one little, tiny pixel. 148 00:06:50.560 --> 00:06:53.800 So the attackers might make these subtle changes, okay, to 149 00:06:54.079 --> 00:06:58.639 the physical layout of the transistors, or introduced tiny little 150 00:06:58.639 --> 00:07:00.160 delays in the way. 151 00:07:00.279 --> 00:07:02.279 That the signals travel through the chip. 152 00:07:02.439 --> 00:07:04.439 So they're messing with the timing exactly. 153 00:07:04.639 --> 00:07:07.759 And these modifications are so small that they can't be 154 00:07:07.839 --> 00:07:09.480 detected through visual inspection. 155 00:07:09.759 --> 00:07:10.680 You can't even see it. 156 00:07:10.720 --> 00:07:13.439 You can't see it. Wow, it's like trying to find 157 00:07:13.480 --> 00:07:16.680 a single misspelling in a library full of books. 158 00:07:16.879 --> 00:07:19.720 That's a good analogy, right, Okay, so that's one technique. 159 00:07:19.759 --> 00:07:20.199 That's one. 160 00:07:20.279 --> 00:07:21.519 What are some other ways? 161 00:07:21.800 --> 00:07:23.600 There are a number of other ways. 162 00:07:23.279 --> 00:07:24.519 That they hide these things. 163 00:07:24.879 --> 00:07:30.319 One clever tactic is to exploit what's called unused or 164 00:07:30.439 --> 00:07:35.319 redundant circuitry. Redundant circuitry, So what is that Modern ships 165 00:07:35.360 --> 00:07:39.319 today they often have these sections of circuitry that are 166 00:07:39.319 --> 00:07:43.360 designed for testing purposes or as backups in case of 167 00:07:43.399 --> 00:07:44.480 a component failure. 168 00:07:44.560 --> 00:07:46.040 It says, like a spare room in a house. 169 00:07:46.279 --> 00:07:47.800 Yeah, exactly, like a spare room. 170 00:07:47.839 --> 00:07:49.040 You don't use it every day. 171 00:07:49.240 --> 00:07:51.319 You don't use it, but it's there. If you need it, 172 00:07:51.399 --> 00:07:55.560 you got it. So attackers they can repurpose, okay, this 173 00:07:55.800 --> 00:07:58.839 unused circuitry for malicious. 174 00:07:58.160 --> 00:08:01.399 Purposes, so they can hide their trojan in these like 175 00:08:01.519 --> 00:08:04.519 inactive areas exactly. So it makes it harder. 176 00:08:04.240 --> 00:08:06.360 To detect, much harder to detect. 177 00:08:06.079 --> 00:08:08.639 Because it's not actually interfering with the normal function of 178 00:08:08.680 --> 00:08:10.560 the chip exactly. So you wouldn't even know. 179 00:08:10.519 --> 00:08:12.319 It's there unless you knew where to look. 180 00:08:12.399 --> 00:08:15.160 So it's like a secret passage hidden behind a bookcase. 181 00:08:15.279 --> 00:08:16.480 That's a great way to put it. 182 00:08:16.560 --> 00:08:18.360 You wouldn't know it was there unless you knew where 183 00:08:18.439 --> 00:08:19.560 to look exactly. 184 00:08:19.800 --> 00:08:23.360 And then there's another technique called logic obfuscation. 185 00:08:23.800 --> 00:08:26.879 Logic obfuscation, This one's a little bit more technical, Okay, 186 00:08:26.920 --> 00:08:27.720 break down for me. 187 00:08:27.959 --> 00:08:34.039 But essentially it involves scrambling the logic of a circuit, okay, 188 00:08:34.159 --> 00:08:37.799 making it very difficult to understand without the right key. 189 00:08:38.039 --> 00:08:39.720 So it's like you're encrypting a message. 190 00:08:39.840 --> 00:08:41.559 Yes, it's like encrypting a message. 191 00:08:41.679 --> 00:08:44.720 You take a readable message, you scramble it with a 192 00:08:44.759 --> 00:08:47.080 secret code exactly, and then you send it. 193 00:08:47.039 --> 00:08:50.120 To someone, right, and only the person who has. 194 00:08:50.080 --> 00:08:52.960 The key with the key can decipher it, can read it. 195 00:08:53.039 --> 00:08:53.440 Got it. 196 00:08:53.919 --> 00:08:56.120 So logic obfuscation does. 197 00:08:55.960 --> 00:09:00.200 Something similar with circuits. Okay. The attacker will take the 198 00:09:00.240 --> 00:09:04.120 malicious code and they will rewrite it in a way 199 00:09:04.120 --> 00:09:07.159 that just looks like gibberish okay to anyone who doesn't 200 00:09:07.200 --> 00:09:07.639 have the key. 201 00:09:07.799 --> 00:09:10.799 So it's like hiding a message in plain sight. Yes, exactly, 202 00:09:10.840 --> 00:09:14.360 the code is there. The code is there, but it's disguised. 203 00:09:14.399 --> 00:09:16.799 It's disguised to look like something else, to look like 204 00:09:16.840 --> 00:09:19.039 something else, so it just blends in with the rest 205 00:09:19.039 --> 00:09:20.080 of the circuitry, is. 206 00:09:20.039 --> 00:09:21.240 Just part of the background noise. 207 00:09:21.399 --> 00:09:24.919 Exactly and it makes it extremely difficult to identify. 208 00:09:25.120 --> 00:09:30.559 Wow, that's so sophisticated. And to make matters worse, they 209 00:09:30.600 --> 00:09:35.559 can have all sorts of crazy activation mechanisms. It's not 210 00:09:35.559 --> 00:09:37.279 like they just switch on as soon as you plug 211 00:09:37.320 --> 00:09:38.159 in your phone charger. 212 00:09:38.559 --> 00:09:43.559 Not necessarily, right, So the activation mechanisms can be very specific. 213 00:09:44.159 --> 00:09:47.840 Some are triggered by a particular date or time, others 214 00:09:47.840 --> 00:09:50.600 by a specific signal or sequence of events. 215 00:09:51.279 --> 00:09:51.840 Wow. 216 00:09:51.879 --> 00:09:55.080 There are even trojans that are designed to activate randomly, so. 217 00:09:55.039 --> 00:09:58.159 They're totally unpredictable, incredibly unpredictable. 218 00:09:58.440 --> 00:10:00.440 So they could be lying dormant free years. 219 00:10:00.720 --> 00:10:03.440 Yeah, potentially just waiting for the right moment to strike. 220 00:10:03.799 --> 00:10:04.720 That is a concern. 221 00:10:05.000 --> 00:10:05.759 That's terrifying. 222 00:10:05.840 --> 00:10:08.480 It is a concern, and this is why detecting hardware 223 00:10:08.480 --> 00:10:11.440 trojans is such a challenge. Yeah, it's not just about 224 00:10:11.480 --> 00:10:14.759 finding a physical piece of hardware, right, but it's about 225 00:10:14.879 --> 00:10:17.360 understanding how it's programmed and. 226 00:10:17.279 --> 00:10:20.159 What triggers it, what might trigger it. Okay, so this 227 00:10:20.200 --> 00:10:23.039 is where the detective work comes in. Yeah, are there 228 00:10:23.080 --> 00:10:28.559 any telltale signs? There are some any clues that researchers 229 00:10:28.559 --> 00:10:28.960 look for. 230 00:10:29.200 --> 00:10:31.159 Yeah, there are some subtle things when. 231 00:10:31.039 --> 00:10:33.120 They're trying to find these potential trojans. 232 00:10:33.320 --> 00:10:38.120 One common red flag is unusual delays in the chips operation. Okay, 233 00:10:38.200 --> 00:10:40.960 so if a chip is taking longer than expected to 234 00:10:41.000 --> 00:10:44.399 perform certain tasks, that could be a sign that a 235 00:10:44.440 --> 00:10:47.200 hidden circuit is interfering with its normal function. 236 00:10:47.679 --> 00:10:49.679 So it's like you're talking to somebody and they have 237 00:10:49.759 --> 00:10:52.559 like a slight hesitation, yes, like a little or dropped 238 00:10:52.600 --> 00:10:56.000 word purple. It might not be obvious at first, but 239 00:10:56.039 --> 00:10:58.919 it could be a sign that something is wrong exactly. 240 00:10:59.159 --> 00:11:03.519 Another sign is unusual power consumption. Okay, So if a 241 00:11:03.600 --> 00:11:06.360 chip is drawing more power than it should. 242 00:11:06.960 --> 00:11:09.360 That could be a sign that something's going on. 243 00:11:09.480 --> 00:11:11.960 That a hidden circuit is active when it shouldn't be. 244 00:11:13.080 --> 00:11:15.919 So it's like looking at your electricity bill, yeah, and 245 00:11:16.000 --> 00:11:18.480 noticing a spike even though you haven't been using any 246 00:11:18.480 --> 00:11:19.279 more appliances. 247 00:11:19.960 --> 00:11:22.399 Exactly. Something's using up that extra power. 248 00:11:22.480 --> 00:11:23.159 Yeah, what is that? 249 00:11:23.399 --> 00:11:26.879 And then there's also these side channel emissions Okay, we 250 00:11:27.000 --> 00:11:31.679 mentioned earlier, things like variations and electromagnetic radiation. These emissions 251 00:11:31.679 --> 00:11:35.200 can actually reveal these subtle patterns of activity within the chip, 252 00:11:35.320 --> 00:11:38.320 and they could potentially exposing the presence of hidden circuitry. 253 00:11:38.399 --> 00:11:41.360 Wow. So it really is like forensic science. 254 00:11:41.399 --> 00:11:42.120 It really is. 255 00:11:42.200 --> 00:11:45.799 You're looking for tiny little clues, tiny little clues that 256 00:11:45.960 --> 00:11:48.600 tell a bigger story exactly. It's a good analogy. 257 00:11:48.720 --> 00:11:53.360 Require specialized equipment, yeah, sophisticated analysis techniques, and a healthy 258 00:11:53.399 --> 00:11:54.240 dose of intuition. 259 00:11:54.639 --> 00:11:57.120 Well, this has been a fascinating if a little terrifying. 260 00:11:57.399 --> 00:11:59.799 Yeah, deep dive thick a complex topic. 261 00:11:59.559 --> 00:12:03.480 Into this world of hardware trojans. But I want to 262 00:12:03.480 --> 00:12:05.879 shift gears a little bit and talk about solutions. 263 00:12:06.000 --> 00:12:06.240 Yeah. 264 00:12:06.759 --> 00:12:10.039 We've talked about some countermeasures, sure, like secure design practices 265 00:12:10.080 --> 00:12:14.440 and trusted manufacturing. Are there any other strategies. 266 00:12:13.840 --> 00:12:14.840 Yes, there are actions. 267 00:12:14.840 --> 00:12:16.240 They are being developed to combat this. 268 00:12:16.519 --> 00:12:18.840 There's a lot of research and development going on in 269 00:12:18.879 --> 00:12:23.240 this area, Okay, some promising advancements, Like what so one 270 00:12:23.360 --> 00:12:27.600 area that's particularly exciting is the development of what's called 271 00:12:27.720 --> 00:12:33.039 specialized hardware security modules. Hardware security hardware security modules. 272 00:12:33.080 --> 00:12:33.600 What are those? 273 00:12:33.840 --> 00:12:38.200 So? These are essentially separate chips that are designed with 274 00:12:38.320 --> 00:12:40.279 security as their primary function. 275 00:12:40.679 --> 00:12:41.840 So they're like a watchdog. 276 00:12:42.039 --> 00:12:44.960 Yes, they act as a sort of watchdog. They're monitoring 277 00:12:45.000 --> 00:12:47.440 the activity of the other chips within a. 278 00:12:47.399 --> 00:12:49.519 System, looking for anything suspicious. 279 00:12:49.799 --> 00:12:53.399 Exactly. So if they detect something fishy, they can take 280 00:12:53.519 --> 00:12:57.039 various actions to prevent or mitigate attacks. 281 00:12:57.200 --> 00:12:59.440 So it's like a security guard station inside your computer. 282 00:12:59.720 --> 00:13:01.399 Yeah, that's a good way to think about it. Just 283 00:13:01.440 --> 00:13:03.399 watching everything, watching everything. 284 00:13:03.000 --> 00:13:06.080 So they can isolate compromised chips. Yeah, they can do that, 285 00:13:06.360 --> 00:13:09.559 prevent unauthorized access to data, or even shut down the 286 00:13:09.720 --> 00:13:10.399 entire system. 287 00:13:10.600 --> 00:13:12.759 Yeah, if an attack is detected. 288 00:13:12.399 --> 00:13:14.200 Oh wow, that's powerful stuff. 289 00:13:14.320 --> 00:13:14.639 It is. 290 00:13:15.200 --> 00:13:16.919 Are there any other promising approaches? 291 00:13:17.200 --> 00:13:19.879 There are? So Another area of active research is called 292 00:13:19.919 --> 00:13:21.120 split manufacturing. 293 00:13:21.639 --> 00:13:23.000 Split manufacturing, what's that? 294 00:13:23.360 --> 00:13:28.200 So this involves dividing up the manufacturing process of a 295 00:13:28.279 --> 00:13:31.159 chip among multiple trusted parties. 296 00:13:31.279 --> 00:13:34.200 So instead of one company having control over the whole 297 00:13:34.200 --> 00:13:37.279 thing exactly, different companies are responsible for different. 298 00:13:37.000 --> 00:13:41.120 Stages, different stages, and each company only has access to 299 00:13:41.159 --> 00:13:44.879 the information, got it, that's necessary for their specific part 300 00:13:44.919 --> 00:13:45.639 of the process. 301 00:13:45.799 --> 00:13:47.000 So they don't see the whole picture. 302 00:13:47.039 --> 00:13:48.519 They don't see the whole picture. So it makes it 303 00:13:48.600 --> 00:13:50.519 much more difficult for an attacker. 304 00:13:50.360 --> 00:13:52.320 So they can't just sneak something in there. 305 00:13:52.200 --> 00:13:54.679 To compromise the chip without being detected. 306 00:13:55.080 --> 00:13:56.080 That's really smart. 307 00:13:56.200 --> 00:14:00.120 It's a clever way to minimize the risk of a 308 00:14:00.360 --> 00:14:01.159 point of failure. 309 00:14:01.279 --> 00:14:03.720 Yeah, you're spreading it out exactly. It seems like there 310 00:14:03.759 --> 00:14:05.679 are a lot of smart people working on this problem. 311 00:14:05.720 --> 00:14:07.519 There are a lot of very smart people. 312 00:14:07.679 --> 00:14:08.399 That's good to hear. 313 00:14:08.559 --> 00:14:12.679 It's a top priority for governments, militaries, and technology companies 314 00:14:12.720 --> 00:14:13.759 all over the world. 315 00:14:14.000 --> 00:14:16.080 Yeah, because the stakes are high. 316 00:14:16.159 --> 00:14:18.320 The stakes are incredibly high, and. 317 00:14:18.320 --> 00:14:20.559 As we talked about the potential consequences, they are. 318 00:14:20.559 --> 00:14:21.840 Really scary to think about. 319 00:14:22.039 --> 00:14:23.080 Yeah, it's not good. 320 00:14:23.240 --> 00:14:26.679 So it's a constant battle. Yeah, a race to stay 321 00:14:26.720 --> 00:14:27.639 ahead of the curve. 322 00:14:28.360 --> 00:14:30.159 So what do you think the future holds. 323 00:14:31.879 --> 00:14:34.399 Well, it's hard to say. 324 00:14:34.200 --> 00:14:35.960 For sure, but with all this investment. 325 00:14:36.120 --> 00:14:40.080 Yeah, with continued investment in research and development and a 326 00:14:40.120 --> 00:14:44.519 commitment to international collaboration, I think we can make significant 327 00:14:44.559 --> 00:14:47.080 strides in mitigating this threat. 328 00:14:47.480 --> 00:14:48.399 Well that's reassuring. 329 00:14:48.600 --> 00:14:50.120 Yeah, it's not all doom and gloom. 330 00:14:50.320 --> 00:14:52.679 Yeah, it sounds like we're not completely helpless. 331 00:14:52.919 --> 00:14:55.600 Right. But before we move on to the final part 332 00:14:55.639 --> 00:14:57.559 of our deep dive, I want to make sure we 333 00:14:57.600 --> 00:15:00.679 give the listeners some key takeaways sure from this part 334 00:15:00.679 --> 00:15:01.399 of the conversation. 335 00:15:02.200 --> 00:15:03.679 So what are the main points? 336 00:15:03.919 --> 00:15:05.200 What do we want them to remember? 337 00:15:06.080 --> 00:15:11.120 Hardware trojans are a real and growing threat. Yeah, they 338 00:15:11.159 --> 00:15:14.600 are not just the stuff of science fiction or spy thrillers. 339 00:15:14.720 --> 00:15:15.960 This is real life. 340 00:15:16.000 --> 00:15:19.639 This is real life, This is happening, This is happening. 341 00:15:19.279 --> 00:15:22.200 And it's a clear and present danger, yes. 342 00:15:22.399 --> 00:15:26.919 To our national security, economic stability, and personal privacy. 343 00:15:27.120 --> 00:15:32.679 And as we've discussed, these trojans can be incredibly difficult. 344 00:15:32.200 --> 00:15:34.440 To detect, very difficult to detect. 345 00:15:34.519 --> 00:15:37.559 They're often hidden in plain sight. They are disguised as 346 00:15:37.679 --> 00:15:40.039 legitimate parts of a ship's circuitry exactly. 347 00:15:40.120 --> 00:15:43.000 So that's why it's so important, Yeah, to be cautious 348 00:15:43.039 --> 00:15:44.600 about the technology we use. 349 00:15:44.720 --> 00:15:47.720 We can't just blindly trust that our devices are secure. 350 00:15:47.879 --> 00:15:49.320 We need to be informed consumers. 351 00:15:49.440 --> 00:15:52.159 Yeah, we need to ask questions, ask questions about security 352 00:15:52.200 --> 00:15:53.240 measures as. 353 00:15:53.000 --> 00:15:55.919 And demand greater transparency from the manufacturers. 354 00:15:56.440 --> 00:16:01.000 Right, And even with all these technological advancements, no silver bullet, 355 00:16:01.799 --> 00:16:04.519 no easy effects, no fool proof solutions. 356 00:16:04.519 --> 00:16:07.080 So security is an ongoing process. 357 00:16:06.639 --> 00:16:08.039 It's a constant evolution. 358 00:16:08.279 --> 00:16:11.000 So we've spent the last two parts of this deep 359 00:16:11.080 --> 00:16:16.159 dive really getting into the nitty gritty of hardware trojans. Yeah, 360 00:16:16.200 --> 00:16:19.799 the detail, how they're made, how they're hidden, the sheer 361 00:16:19.879 --> 00:16:20.919 scale of the threat. 362 00:16:21.159 --> 00:16:22.279 Right, it's a lot. 363 00:16:22.519 --> 00:16:24.320 It's a lot to take in, it really is. But 364 00:16:24.440 --> 00:16:28.679 now I'm curious about what the future holds, what's coming 365 00:16:28.759 --> 00:16:34.000 next in this constant battle between those who want to 366 00:16:34.039 --> 00:16:37.200 exploit these vulnerabilities and those who are trying to defend 367 00:16:37.200 --> 00:16:37.679 against them. 368 00:16:37.720 --> 00:16:39.840 Well, you're right, it is this constant back and forth, 369 00:16:39.879 --> 00:16:43.279 and honestly, predicting the future of cyber warfare it's kind 370 00:16:43.279 --> 00:16:46.279 of like trying to predict the weather. Things change so quickly. 371 00:16:46.879 --> 00:16:51.240 But there are definitely some key trends and emerging technologies 372 00:16:51.279 --> 00:16:53.799 that are worth keeping a close eye on. Like one 373 00:16:53.799 --> 00:16:58.919 that's particularly concerning is the rise of artificial intelligence. Yeah 374 00:16:59.159 --> 00:17:03.679 a yeah, everyone's talking about AI these days, it seems. 375 00:17:03.320 --> 00:17:06.039 Like it, but I haven't really thought about it in 376 00:17:06.079 --> 00:17:07.640 the context of cybersecurity. 377 00:17:07.799 --> 00:17:10.519 Yeah, well think about it this way. AI can be 378 00:17:10.680 --> 00:17:14.319 used to automate a lot of these really complex tasks, 379 00:17:14.839 --> 00:17:17.680 and that includes tasks related to cyberwarfare. 380 00:17:18.119 --> 00:17:18.319 Oh. 381 00:17:18.720 --> 00:17:22.160 So on the offensive side, attackers could use AI to 382 00:17:22.279 --> 00:17:28.480 identify vulnerabilities, develop much more sophisticated malware, and even launch 383 00:17:28.519 --> 00:17:31.559 attacks that adapt to changing defenses in real time. 384 00:17:31.640 --> 00:17:34.960 So it's like giving the bad guys a superpowered toolkit. 385 00:17:35.160 --> 00:17:37.319 That's a good way to think about it. Okay, But 386 00:17:37.400 --> 00:17:39.039 there's a flip side to this as well. 387 00:17:39.200 --> 00:17:39.440 Okay. 388 00:17:39.599 --> 00:17:42.240 Security researchers can also leverage. 389 00:17:41.880 --> 00:17:44.720 AI right for good for good, Okay. 390 00:17:44.960 --> 00:17:48.319 AI can be used to analyze massive amounts of data, 391 00:17:49.160 --> 00:17:52.000 identify patterns of malicious activity. 392 00:17:51.680 --> 00:17:52.799 So you can see it coming. 393 00:17:52.759 --> 00:17:56.759 Yeah, and develop countermeasures much more quickly than humans could 394 00:17:56.839 --> 00:17:57.279 on their own. 395 00:17:58.319 --> 00:18:01.839 So it's an arms race. Both sides are using AI. 396 00:18:02.079 --> 00:18:03.599 Absolutely, it's a constant race. 397 00:18:03.640 --> 00:18:05.200 Wow. Okay, so that's AI. 398 00:18:06.000 --> 00:18:11.920 What about another trend that's worth watching is this increasing 399 00:18:12.160 --> 00:18:16.440 integration of technology into every aspect of our lives. This 400 00:18:16.599 --> 00:18:19.440 Internet of things right where everything is connected, Where everything 401 00:18:19.480 --> 00:18:23.559