TRANSKRIPT 🔗
WEBVTT
1
00:00:00.231 --> 00:00:01.953
I didn't understand the context of this.
2
00:00:02.053 --> 00:00:05.636
I thought this was a guy, like, detecting that he was getting spied on.
3
00:00:05.956 --> 00:00:16.566
I did not realize that this picture was the spy committing the espionage, seeing that he was getting counter-espionage, getting tracked, and it's him detecting the plant.
4
00:00:16.586 --> 00:00:18.948
He's like, oh, no, I'm f***ed.
5
00:00:25.113 --> 00:00:27.235
Good morning, hackers.
6
00:00:27.255 --> 00:00:31.439
Welcome back to the lowdown, the best podcast on the internet for hackers.
7
00:00:31.479 --> 00:00:33.240
Matt, dude, it's been a week.
8
00:00:33.741 --> 00:00:35.162
Life got in the way last week.
9
00:00:35.182 --> 00:00:43.149
We were not able to record, but as a shock to nobody, things still happen in cyberspace despite our absence, and some of them are pretty crazy, actually.
10
00:00:43.189 --> 00:00:46.472
But before we get into that, I saw you gave a keynote recently.
11
00:00:46.832 --> 00:00:47.213
How was that?
12
00:00:48.128 --> 00:00:49.729
I'm going to redo my pose here.
13
00:00:50.110 --> 00:00:53.733
Yeah, I was like pointing at the thing with the logo in the back.
14
00:00:53.793 --> 00:00:54.514
Come on, it's a good shot.
15
00:00:54.534 --> 00:01:01.139
Josh, please take the photo on LinkedIn, invert it, and have Matt do a finger-to-finger touch a la Da Vinci.
16
00:01:01.279 --> 00:01:01.680
Please.
17
00:01:01.820 --> 00:01:02.881
Oh, it's perfect.
18
00:01:02.901 --> 00:01:04.122
Let me get my finger on the screen then.
19
00:01:04.842 --> 00:01:06.003
Yes, let's go.
20
00:01:07.264 --> 00:01:09.305
Yeah, so Sektor is cool.
21
00:01:09.525 --> 00:01:11.525
Black Hat Canada, straight up.
22
00:01:11.785 --> 00:01:15.786
Definitely less hacker vibes than Vegas, just FYI, if you ever go, right?
23
00:01:16.687 --> 00:01:18.607
But the location's great.
24
00:01:18.627 --> 00:01:20.108
You're right there in downtown Toronto.
25
00:01:20.428 --> 00:01:28.710
All the buildings around you are all the buildings of like the biggest companies in Canada and all the people at the conference are the security teams from those companies.
26
00:01:29.670 --> 00:01:43.117
Got to meet, you know, security practitioners and leadership from Canada's biggest telecoms, energy, you know, power companies, insurance and banking sector is really big right there.
27
00:01:43.217 --> 00:01:44.898
It's kind of right in financial district.
28
00:01:45.218 --> 00:01:45.358
Nice.
29
00:01:45.378 --> 00:01:50.080
And so, yeah, I got to do the opening keynote for the AI Summit before the briefing started.
30
00:01:50.120 --> 00:01:52.582
So it was just pure play AI day, right?
31
00:01:52.602 --> 00:01:52.722
Yeah.
32
00:01:53.782 --> 00:01:54.282
It was kind of funny.
33
00:01:54.322 --> 00:01:55.523
I made this joke in my talk.
34
00:01:55.963 --> 00:02:00.465
I made it about 10 minutes into this 25 minute talk when I started
Load full transcript
talking about the hugging face incident.
35
00:02:00.965 --> 00:02:03.287
And I was like, oh, put a dollar in the jar and hey, everyone clap.
36
00:02:03.307 --> 00:02:05.628
I made it 10 minutes without saying hugging face.
37
00:02:05.648 --> 00:02:06.908
I crap you not.
38
00:02:07.869 --> 00:02:10.670
Every talk after me did not make it 60 seconds.
39
00:02:11.030 --> 00:02:11.310
Really?
40
00:02:11.370 --> 00:02:12.711
Like, yeah, yeah, yeah.
41
00:02:12.751 --> 00:02:13.872
Like, I'm not even dunking on it.
42
00:02:13.952 --> 00:02:19.435
It was just so funny that, like, I sat there the rest of the day and I laughed because I made it 10 minutes and I kind of rolled my eyes at myself.
43
00:02:19.775 --> 00:02:21.356
No one else made it 90 seconds.
44
00:02:21.556 --> 00:02:22.817
Let's be generous, right?
45
00:02:23.477 --> 00:02:24.078
So, yeah.
46
00:02:24.258 --> 00:02:27.640
Topic du jour was the AI super hackers are here.
47
00:02:27.720 --> 00:02:28.801
What are we doing about it?
48
00:02:29.841 --> 00:02:31.162
I will say, I will say.
49
00:02:31.322 --> 00:02:53.778
And this is, again, not to like dunk on anyone that I met or talked to, but I will say the more I go like out into the real world and conferences and stuff and not like my little bubbles of group chats and things like that and talking to you and whatever, the gap of understanding the current capabilities as it stands with AI is wide, even at the highest levels of our industry.
50
00:02:54.098 --> 00:02:54.318
Right.
51
00:02:54.718 --> 00:03:01.122
So like I said, I'm talking to leadership that are running security departments for large financial energy, whatever institutions.
52
00:03:02.063 --> 00:03:10.487
And you can kind of tell that some people made up their mind about AI's capabilities way too long ago, way too long ago.
53
00:03:10.708 --> 00:03:12.869
Kind of the phase you were in, right?
54
00:03:13.369 --> 00:03:14.530
You were one of these people, right?
55
00:03:15.210 --> 00:03:19.634
Um, but to your credit, I think you did what a lot of people in our industry should be doing.
56
00:03:19.774 --> 00:03:23.577
You stay hands on with this stuff because it changes at such light speed.
57
00:03:23.598 --> 00:03:24.098
Right.
58
00:03:24.178 --> 00:03:24.398
Right.
59
00:03:24.979 --> 00:03:26.520
Um, that, that, yeah.
60
00:03:26.560 --> 00:03:32.065
So like, that's, that's what I'm coming away from that with is encouraging anyone who's listening.
61
00:03:33.394 --> 00:03:39.079
Don't like just have your opinion and be like the AI hater or the AI obsessed.
62
00:03:39.459 --> 00:03:43.062
Like there's like a nuance in the middle of like what's really going on.
63
00:03:43.082 --> 00:03:44.363
What are things really good at?
64
00:03:44.844 --> 00:03:47.466
And I think I'm encouraging people find that out for yourself.
65
00:03:47.666 --> 00:03:49.848
Like do your own experimentation, stuff like that.
66
00:03:50.368 --> 00:03:51.089
It's super important.
67
00:03:51.369 --> 00:03:52.630
Yeah, I found that to be the case.
68
00:03:52.670 --> 00:04:00.115
Now, again, I came from more of the offensive security consulting world, less the actual CISO in a company world.
69
00:04:00.636 --> 00:04:12.524
But I think I've always found that to be the same, where it's people in these higher-up positions, not that they don't understand, it's just the level and depth of understanding is typically a lot more thin, I guess.
70
00:04:13.024 --> 00:04:16.967
And then on top of that, they develop opinions early, because they have to, it's the nature of their job.
71
00:04:17.368 --> 00:04:18.488
But then because of the...
72
00:04:19.449 --> 00:04:46.165
position they're in the reputation they have to maintain they don't want to look like a flop and like change their opinion a lot so once they kind of like dig into these early opinions and like you said admittedly i did that for a while i was like no ai is stupid ai can't hack but again i don't work for a fortune 500 company anymore uh and so when i decided i want to change it you know i played around yeah man the models are getting really kind of scary good as we'll go into here in the the rest of the podcast at hacking it's it's pretty bonkers yeah
73
00:04:46.420 --> 00:04:52.144
So man, you're putting out banger content on your channel, literally about how good this stuff is at reversing right now.
74
00:04:52.985 --> 00:04:53.205
Yeah.
75
00:04:53.265 --> 00:04:58.669
Not only podcast co-host number one, low-level fan over here, but you know, you're like this stuff is getting scary.
76
00:04:58.689 --> 00:04:58.930
Good.
77
00:05:00.031 --> 00:05:08.317
The others, like I got pulled into a few other talks because people started to realize that I was in Canada and they were like, well, we can't let you just do 25 minutes on stage.
78
00:05:08.337 --> 00:05:12.160
So they pulled me up for kind of a second panel and then a third panel during the briefings.
79
00:05:12.360 --> 00:05:12.880
Nice.
80
00:05:12.980 --> 00:05:14.882
And it does not matter where I was.
81
00:05:14.922 --> 00:05:16.743
This is all anyone was talking about.
82
00:05:16.883 --> 00:05:17.323
This is it.
83
00:05:17.483 --> 00:05:19.445
Like, this is all anyone wanted to talk to me about.
84
00:05:19.465 --> 00:05:21.666
This is all anyone wanted to, like, whatever.
85
00:05:22.206 --> 00:05:24.027
Like, this was it.
86
00:05:24.087 --> 00:05:30.852
So even though it was cute that we had this little AI summit, it might as well have been the whole thing, man, because everyone's just trying to figure it out right now.
87
00:05:31.479 --> 00:05:32.320
Yeah, I mean, it is weird.
88
00:05:32.380 --> 00:05:39.947
Again, I'm not going to try to regurgitate the entire thesis of your talk, but it also is part of the video that I made recently that I want to highlight.
89
00:05:40.308 --> 00:05:46.634
The reason why I think everyone's talking about it is that it truly has commoditized being a nation
state, right?
90
00:05:46.654 --> 00:05:50.557
Name of the channel, by the way, where, again, man, I made a video on this.
91
00:05:50.577 --> 00:05:56.183
A task that used to take me a week plus is now a 15-minute side quest.
92
00:05:56.583 --> 00:05:57.863
right, using GLM-5.3.
93
00:05:57.883 --> 00:06:08.046
For people that don't know, I made a video, right, like, reverse-engineered, the firmware of my monitor, right, upstairs in my office in my other part of my house, it ships encrypted.
94
00:06:08.386 --> 00:06:11.727
Now, it also ships with the DLL that does the decryption.
95
00:06:12.267 --> 00:06:21.129
Now, obviously, the key is in there, but the process of reverse-engineering, like, the structure and the format and, like, where the key material is and how it's decrypted and encrypted, you can do that.
96
00:06:21.229 --> 00:06:22.469
Anyone can do that if you know how to RE.
97
00:06:22.509 --> 00:06:23.269
It just takes a minute.
98
00:06:23.349 --> 00:06:26.050
But again, now the AIs can just do it for you lickety-split,
99
00:06:26.770 --> 00:06:27.691
It's really bonkers.
100
00:06:28.191 --> 00:06:28.471
Yeah.
101
00:06:28.891 --> 00:06:44.077
You know, there was just a few times where I was like hearing people literally on stage talking and saying, well, yeah, now you got to deal with all these like AI slop code, vibe coded apps filled with vulnerabilities.
102
00:06:45.218 --> 00:06:46.398
And like, I get it.
103
00:06:46.558 --> 00:06:47.559
Like, I do.
104
00:06:47.719 --> 00:06:49.700
I like really do get like it.
105
00:06:51.672 --> 00:06:55.914
Not to be too AI-pilled, but vulnerabilities have been around a long time.
106
00:06:56.014 --> 00:06:56.795
Right.
107
00:06:57.995 --> 00:07:04.418
It's not like we were putting out pristine, vuln-free code, organic, handcrafted.
108
00:07:04.759 --> 00:07:05.819
Artisanal, yeah.
109
00:07:06.440 --> 00:07:06.760
No.
110
00:07:07.000 --> 00:07:11.002
And, like, have you used Astra 6 or Opus 5.5?
111
00:07:11.142 --> 00:07:16.885
And, like, are you throwing out SQL injection or cross-site scripting vulnerabilities with those things?
112
00:07:17.646 --> 00:07:21.008
No, I'm not saying we've solved security and software is solved or anything like that.
113
00:07:21.248 --> 00:07:34.574
But you can't make up your mind on chatGPT.com version 3 or whatever the heck it was however many years ago and be like, well, look at all the hallucinations and all the stuff that is going on and be like, see, that's it.
114
00:07:34.674 --> 00:07:36.175
Look, I told you it's so bad.
115
00:07:36.635 --> 00:07:41.498
Let me give you a concrete example, you, the audience, of I think how...
116
00:07:42.438 --> 00:07:50.445
different things are in terms of, you know, AI code is insecure versus AI code is safe or more secure.
117
00:07:50.645 --> 00:07:50.845
Right.
118
00:07:51.166 --> 00:07:56.750
So again, I've been using the AI models generate code since like the three, three, five days where I did the primogens tower event.
119
00:07:56.770 --> 00:08:01.574
And when we did it back then, the models created terrible code, insecure, not maintainable, whatever.
120
00:08:02.455 --> 00:08:10.120
But if you look at the models today and you look at an incident like the Citrix net scaler bugs, and I'm going to do a lot of dunking on Citrix here in the next couple of seconds.
121
00:08:10.140 --> 00:08:11.921
I'm not trying to like drag them through the mud.
122
00:08:12.161 --> 00:08:16.444
This is just a very, I think, poignant example of how things could be better, right?
123
00:08:17.004 --> 00:08:27.490
A lot of the vulnerabilities in those platforms were literally just like the N value for a field came off the network and was not validated before a mem copy occurred, right?
124
00:08:27.510 --> 00:08:29.592
That's like traditional buffer overflow type stuff.
125
00:08:30.212 --> 00:08:32.273
The models do not produce that pattern.
126
00:08:32.613 --> 00:08:33.834
They literally don't anymore.
127
00:08:33.954 --> 00:08:45.118
If you were to vibe code, what is it, NSA, AAD, the daemon that does the AAA authentication on the NetScaler, it would not produce that vulnerable path anymore.
128
00:08:45.158 --> 00:08:53.362
So we're starting to see a shift where sometimes, at least based on certain companies and their coding practices, AI code may be more secure.
129
00:08:53.602 --> 00:08:54.022
You know what I mean?
130
00:08:54.042 --> 00:08:56.023
They're just not capable, not capable in quotes.
131
00:08:56.043 --> 00:08:56.864
We catch a lot of shit for that.
132
00:08:57.124 --> 00:09:00.646
They statistically produce less that code pattern.
133
00:09:00.866 --> 00:09:01.407
You know what I mean?
134
00:09:01.427 --> 00:09:02.627
It's just, it's really interesting.
135
00:09:03.007 --> 00:09:03.228
Yeah.
136
00:09:03.268 --> 00:09:09.591
Especially on the, like the injection code quality, memory safety, all the, like those syntactical vulnerabilities, right?
137
00:09:09.631 --> 00:09:12.713
Like we're obviously going to start seeing less and less of that as the models get better.
138
00:09:13.233 --> 00:09:13.954
Don't get me wrong.
139
00:09:14.254 --> 00:09:20.758
I'm still seeing and still running into like in person on projects that I'm working on more business logic type stuff.
140
00:09:20.838 --> 00:09:21.038
Right.
141
00:09:21.258 --> 00:09:24.281
more authorization type things, right?
142
00:09:24.401 --> 00:09:27.724
More, oh, we didn't actually bound this the right way.
143
00:09:28.705 --> 00:09:31.027
It should only be accessed from this person, whatever, right?
144
00:09:31.067 --> 00:09:32.869
That kind of stuff could still happen for sure.
145
00:09:32.889 --> 00:09:35.492
We're not up here being like, hey, problem solved.
146
00:09:35.512 --> 00:09:36.312
Everyone go home.
147
00:09:36.933 --> 00:09:44.120
But to stand on stage and be like this AI slop code, I'm like, what have we been doing for 20 years if not slop
coding things?
148
00:09:45.267 --> 00:09:46.809
with our little fingers here.
149
00:09:46.829 --> 00:09:47.569
You know what I mean?
150
00:09:47.589 --> 00:09:50.672
Like I've written a lot of human slop code.
151
00:09:50.692 --> 00:09:51.353
I'll tell you that.
152
00:09:52.454 --> 00:09:55.697
So anyway, you know, that's what I, when I say the gap, that's what I mean.
153
00:09:55.897 --> 00:09:56.117
Right.
154
00:09:56.177 --> 00:10:02.263
It's just like, you can kind of tell the people that have like stayed on top of it and are using this stuff.
155
00:10:02.843 --> 00:10:05.485
And especially, especially, especially with like,
156
00:10:06.246 --> 00:10:09.249
modern harnesses, and things like that, right?
157
00:10:09.269 --> 00:10:24.665
Like, I always pull up when that Mythos thing happened, Project Glasswing thing happened, the companies that were really transparent about their experience with it, the Cloudflare's, the Mozilla's, the Google's, like, all this kind of stuff, and they start writing blogs about their experience with it, it's like, oh, you could tell.
158
00:10:25.025 --> 00:10:25.986
You could tell that these are
159
00:10:26.827 --> 00:10:34.650
top, top, top tier security departments because they're like, look, I just got a better tool to do a lot of the stuff that I was doing.
160
00:10:34.970 --> 00:10:37.151
But look at this crazy harness that I put it into.
161
00:10:37.472 --> 00:10:41.413
I already had this like automation flow and I just injected steroids into it.
162
00:10:41.853 --> 00:10:42.554
Ta-da-da-da.
163
00:10:42.594 --> 00:10:49.617
Look at what happens on the other side versus people who didn't have that to begin with and are looking at AI to solve all their problems.
164
00:10:49.677 --> 00:10:52.398
And they're not, it's not magically solving all their problems.
165
00:10:52.518 --> 00:10:52.718
Yeah.
166
00:10:53.098 --> 00:10:55.280
And they're like, well, I don't get it.
167
00:10:55.400 --> 00:10:56.702
I don't get how it's so much better.
168
00:10:56.762 --> 00:11:01.546
It's like, well, that means you probably weren't doing the right security engineering thing to begin with.
169
00:11:01.806 --> 00:11:03.248
Like, need to break it to you, right?
170
00:11:04.949 --> 00:11:06.431
So anyway, Soapbox, right?
171
00:11:06.591 --> 00:11:17.882
It's just like, the more I go into the real world, the more I realize the gap between like the tip of the spear security programs and like people that are kind of reacting to headlines and doing things, you know, the way that they've always done them.
172
00:11:18.760 --> 00:11:30.649
Yeah, it just boils down to, like, you still have to have a fundamental understanding of, like, what is required to publish secure code to maintain a secure system to have a secure network, right?
173
00:11:30.669 --> 00:11:33.731
Like, AI cannot rid you of those requirements, right?
174
00:11:33.751 --> 00:11:37.955
I think a lot of companies may be taking the AI pill too far.
175
00:11:37.975 --> 00:11:40.457
Again, we're going to come off as AI pilled in this podcast.
176
00:11:40.557 --> 00:11:41.057
I'm not.
177
00:11:41.457 --> 00:11:47.282
It's just you have to acknowledge the benefits and the downsides that the technology offers, right?
178
00:11:47.302 --> 00:11:47.402
Yeah.
179
00:11:47.762 --> 00:11:53.390
All I'm asking, I'm not like the super AI pill that everything is magic land either.
180
00:11:53.631 --> 00:12:00.100
All I'm asking is for if you're listening and you have your mind made up and you haven't touched this stuff in a while,
181
00:12:01.440 --> 00:12:05.224
Like, go do some experiments, go see what's going on, try it out.
182
00:12:05.504 --> 00:12:18.236
This is the only way, I think security people, especially security people, it's more important for us to keep our hands on this stuff as it evolves, to really understand it, to really understand its capabilities, because our devs are going to do it anyway.
183
00:12:18.536 --> 00:12:19.678
They're going to do it no matter what.
184
00:12:19.998 --> 00:12:22.400
They're going to be using the new tool, right?
185
00:12:22.440 --> 00:12:25.183
So we have to understand what it's capable of and where we're at.
186
00:12:25.443 --> 00:12:26.203
Yeah, and it moves so quick.
187
00:12:26.223 --> 00:12:29.545
I mean, you said back when Glasswing happened, bro, that was like five months ago.
188
00:12:29.585 --> 00:12:30.206
You know what I mean?
189
00:12:30.306 --> 00:12:33.948
Like, that was not that long ago, but it feels that long ago, because that's how fast this industry moves.
190
00:12:34.328 --> 00:12:35.449
Right.
191
00:12:35.469 --> 00:12:36.710
You know what else moves quickly, Matt?
192
00:12:37.762 --> 00:12:48.090
My heart rate when I saw that Ledger crypto wallets were getting drained because that's where I may or may not store my troves of 10th Bitcoin.
193
00:12:49.371 --> 00:12:50.812
Yeah, man, it's pretty crazy.
194
00:12:50.852 --> 00:12:58.538
So if you guys aren't following, there has been a series of backdoors, both in software and physical backdoors.
195
00:12:59.318 --> 00:13:00.859
put into crypto wallets.
196
00:13:01.360 --> 00:13:02.501
A news headline came up.
197
00:13:02.661 --> 00:13:10.166
I think it started on Twitter where someone reported that, hey, ledger wallets around the world are just getting drained to zero and no one really knows why.
198
00:13:11.107 --> 00:13:15.450
And then a couple days later, a guy came out and he's like, hey, does this look normal to you guys?
199
00:13:15.490 --> 00:13:17.832
And it was like a physical implant
200
00:13:18.332 --> 00:13:28.258
In a wallet that supposedly was sucking out the seed phrase when it got initialized, when you first plug it in and create your wallet.
201
00:13:28.498 --> 00:13:30.699
So, you know, talk about supply chain attack, man.
202
00:13:30.719 --> 00:13:32.960
We talk a lot about, like, NPM worm type shit.
203
00:13:32.980 --> 00:13:35.942
This is like a real-life hardware supply chain attack.
204
00:13:36.122 --> 00:13:41.486
You know what came to mind for me right away, and it's a little dark, and so I apologize, is those pagers.
205
00:13:41.866 --> 00:13:45.389
Yeah, I wasn't going to say it, but it's floating around in my head.
206
00:13:45.449 --> 00:13:45.689
Yeah.
207
00:13:45.729 --> 00:13:51.173
Well, I mean, it's just like the similarity of the attack itself, right?
208
00:13:51.273 --> 00:13:57.177
Like how upstream the threat actors went in order to pull off this attack.
209
00:13:57.357 --> 00:13:59.118
Obviously, that was a military operation.
210
00:13:59.138 --> 00:14:01.059
If you don't know, we're not going to get into it.
211
00:14:01.440 --> 00:14:02.621
It's not that kind of podcast.
212
00:14:02.681 --> 00:14:03.241
Yeah.
213
00:14:04.502 --> 00:14:06.585
military operation pager supply chain.
214
00:14:06.845 --> 00:14:07.646
Go have fun reading.
215
00:14:07.666 --> 00:14:09.147
Google.
216
00:14:09.167 --> 00:14:09.648
Go have fun.
217
00:14:09.768 --> 00:14:10.649
We're not going to talk about it.
218
00:14:11.029 --> 00:14:18.017
But the similarities here are like, oh, we went and acquired a company and
219
00:14:18.637 --> 00:14:31.486
acquired a company to do resale of hardware because then we could intercept these shipments, leech off the trust that this reseller has built being a legitimate reseller for as long as they have.
220
00:14:31.506 --> 00:14:31.986
What was it called?
221
00:14:32.006 --> 00:14:32.727
Crypto bills?
222
00:14:33.067 --> 00:14:33.267
Yes.
223
00:14:33.307 --> 00:14:33.688
I believe.
224
00:14:34.048 --> 00:14:34.248
Yep.
225
00:14:34.628 --> 00:14:41.055
And so they were a legitimate, so ledgers, legitimate, crypto bills, legitimate until it wasn't.
226
00:14:41.356 --> 00:14:41.616
Right.
227
00:14:41.696 --> 00:14:48.683
All of a sudden someone came and acquired crypto bills and was like, okay, now we are going to ride the reputation of this company.
228
00:14:49.024 --> 00:14:53.188
We have all of a sudden a warehouse full of hardware that then we can inject.
229
00:14:53.208 --> 00:14:53.549
Right.
230
00:14:54.447 --> 00:14:56.089
are attacking to.
231
00:14:56.509 --> 00:14:56.689
Right?
232
00:14:56.749 --> 00:15:07.260
And the other kind of interesting part here to me was, well, there's like a side story here we could talk about about non-ledger, other wallets also having some of these similar issues.
233
00:15:07.700 --> 00:15:11.143
But on the ledger side, it's...
234
00:15:12.705 --> 00:15:14.226
they didn't mess with the hardware itself.
235
00:15:14.586 --> 00:15:20.972
And that's part of the draw of a hardware wallet is like, oh, look, it's more secure than like a website.
236
00:15:21.072 --> 00:15:27.498
If your account gets compromised on like some website, then your crypto is just gone, which we see constantly right now, right?
237
00:15:27.618 --> 00:15:28.899
Social engineers, especially.
238
00:15:30.280
--> 00:15:34.282
So the hardware wallet is supposed to be a little bit more secure, a little bit more tamper-proof.
239
00:15:34.302 --> 00:15:44.947
So if you actually tampered with the hardware itself from Ledger for the actual wallet, that should be detectable or not possible, right?
240
00:15:45.368 --> 00:15:48.649
Depending on some of the different attacks that would actually go on against this hardware.
241
00:15:48.669 --> 00:15:49.570
But they didn't do that.
242
00:15:50.050 --> 00:15:53.152
They soldered, they just like used the motherboard that was there to,
243
00:15:53.792 --> 00:15:55.514
And soldered their implant.
244
00:15:56.034 --> 00:16:04.561
And all their implant did was intercept communications between the actual ledger hardware and the monitor.
245
00:16:05.322 --> 00:16:07.924
Just that little display screen that came on the wallet.
246
00:16:08.405 --> 00:16:12.468
And then guess what the display screen shows while you're setting up your ledger wallet?
247
00:16:12.728 --> 00:16:13.869
The recovery key.
248
00:16:13.889 --> 00:16:15.069
The secret phrase, right?
249
00:16:15.089 --> 00:16:15.350
Yeah.
250
00:16:15.450 --> 00:16:21.773
And then the other thing that I thought was really cool here and clever was that they got sell into it in a very, very, very small package.
251
00:16:21.813 --> 00:16:24.455
You'll notice in this image, too, like there's the eSIM right there.
252
00:16:24.515 --> 00:16:27.056
And also they use solder as an antenna.
253
00:16:27.357 --> 00:16:27.517
Yeah.
254
00:16:27.537 --> 00:16:29.178
Like that thing is an antenna floating in there.
255
00:16:29.238 --> 00:16:29.858
It's insane.
256
00:16:30.418 --> 00:16:31.299
smart engineering, right?
257
00:16:31.319 --> 00:16:32.981
So yeah, the SIM is super small.
258
00:16:33.301 --> 00:16:36.625
I saw one of the examples, the SIM itself was like two millimeters.
259
00:16:37.105 --> 00:16:44.292
But then obviously there's like an antenna and some other stuff going on to actually wind up getting cell service to phone home.
260
00:16:44.713 --> 00:16:48.436
So we've got over $100 million in losses.
261
00:16:48.476 --> 00:16:50.638
It was 86 when I first made a video about this.
262
00:16:50.658 --> 00:16:51.159
Yeah.
263
00:16:51.639 --> 00:16:55.482
Over $100 million in losses now, 500 plus addresses.
264
00:16:55.822 --> 00:17:01.546
A lot of this stuff was in, it seemed, Southeast Asia to start, which is where I think that reseller was active.
265
00:17:01.586 --> 00:17:09.251
But then even today, this morning, I saw someone walk into a German store retailer, like on the shelf store.
266
00:17:09.331 --> 00:17:15.939
box on the shelf at a legitimate retailer covered in like anti-theft boxes, right?
267
00:17:15.999 --> 00:17:20.904
Like at a retailer, pulled that off, opened it up and found the implant in there too.
268
00:17:20.924 --> 00:17:25.810
So these things, I feel like we're going to be hearing about this for quite a while.
269
00:17:25.830 --> 00:17:27.552
It's really hard to walk this back, right?
270
00:17:27.832 --> 00:17:30.194
Yeah, so I want to highlight, like, first of all, I'm not a crypto bro.
271
00:17:30.214 --> 00:17:32.095
I'm about to sound like one, but I'm not.
272
00:17:33.016 --> 00:17:38.760
The phrase, not your keys, not your crypto, is what kind of sets the stage here, right?
273
00:17:39.180 --> 00:17:45.885
So the reason why people use hardware wallets is the idea is that, like, if you have your keys on, if you have your crypto on Coinbase, for example,
274
00:17:46.685 --> 00:17:57.548
The fact that your crypto is on Coinbase means that Coinbase also has access to your private key and public key because Coinbase has to be able to issue transactions, right?
275
00:17:57.568 --> 00:18:00.209
It has to be able to sign transactions and put them on the blockchain.
276
00:18:00.229 --> 00:18:09.731
So the idea is, like, if Coinbase either gets a government warrant or, like, Coinbase gets compromised, you know, the keys are there, and therefore they can move your crypto around without your consent.
277
00:18:09.751 --> 00:18:12.292
So therefore, by migrating your crypto into a hardware wallet,
278
00:18:13.095 --> 00:18:15.416
you know, it makes it more secure, which is still the case, by the way.
279
00:18:15.436 --> 00:18:16.217
I still stand by that.
280
00:18:16.237 --> 00:18:20.639
Like you'd be using a hardware security module on the inside of a wallet to store your key.
281
00:18:20.920 --> 00:18:29.024
And then keeping your crypto in that wallet is significantly more secure than, you know, any kind of on a market type storage.
282
00:18:29.585 --> 00:18:33.967
But that being said, there is an inherent trust you put there with the quality of the hardware.
283
00:18:33.987 --> 00:18:38.230
And if it's been physically implanted to extract your seed phrase at boot,
284
00:18:38.810 --> 00:18:40.091
You're going to have a bad time, right?
285
00:18:40.471 --> 00:18:42.792
Now, the question, I guess, is like, what do you do about this?
286
00:18:43.812 --> 00:18:45.373
Not going to lie, it's a very hard problem, right?
287
00:18:45.393 --> 00:18:56.458
Because not everyone has the tooling, the knowledge, the resources to open up their crypto wallet and examine it for hardware implants, right?
288
00:18:56.478 --> 00:19:03.622
So I'm not really sure what the average person does here, but it does kind of just highlight the inherent vulnerability of even hardware supply chain trust.
289
00:19:04.193 --> 00:19:07.715
Yeah, I think buying direct is like the advice that a lot of people are giving, right?
290
00:19:07.835 --> 00:19:11.057
Is like, don't go to a retailer or a reseller at all for some of this hardware.
291
00:19:11.077 --> 00:19:12.457
Kind of helps you out.
292
00:19:12.958 --> 00:19:16.600
But also credit to MG of OMG Cable fame.
293
00:19:17.100 --> 00:19:20.542
He's got a really great thread over on Twitter going through all the different
hardware.
294
00:19:20.642 --> 00:19:22.403
And then also, he's got a crystal ball.
295
00:19:23.048 --> 00:19:27.550
I didn't realize he's got a tweet from the year 2021.
296
00:19:27.570 --> 00:19:28.950
Decades ago.
297
00:19:29.270 --> 00:19:32.091
The year was 2021 in July.
298
00:19:33.192 --> 00:19:37.533
And he has a tweet talking about some research he did.
299
00:19:38.173 --> 00:19:40.614
If you guys don't know, OMG cables.
300
00:19:40.814 --> 00:19:42.895
These are like the hardware hacking cables.
301
00:19:43.135 --> 00:19:45.896
He makes a lot of hardware pen testing type.
302
00:19:46.856 --> 00:19:48.378
So this is his specialty.
303
00:19:48.778 --> 00:19:59.006
And yeah, he was talking about Bitcoin stealing ledger implants that are now invisible and don't have a lot of this stuff that we're finding in some of these pictures, even from back then.
304
00:19:59.206 --> 00:20:03.470
So there are even stealthier attacks possible on this kind of stuff.
305
00:20:03.750 --> 00:20:05.451
Going into even stealthier attacks, right?
306
00:20:05.471 --> 00:20:08.733
So what we're talking about now is the Ledger wallet, right?
307
00:20:08.753 --> 00:20:12.155
The Ledger wallet attack has a physical implant, right?
308
00:20:12.175 --> 00:20:18.639
You open up your Ledger wallet, you'll see, like, oh, this other chip that has an eSIM, a cellular module, and, like, an antenna, right?
309
00:20:18.659 --> 00:20:19.260
Yeah.
310
00:20:19.820 --> 00:20:30.589
Later, we also found, not we, but the world, Trezor wallets were also compromised, but not by a physical implant, by a software implant.
311
00:20:31.049 --> 00:20:32.030
Now, how would you even do this?
312
00:20:32.050 --> 00:20:35.133
How would you pull off a software compromise of the Trezor wallets?
313
00:20:35.153 --> 00:20:37.255
If you don't remember, Joe Grand...
314
00:20:37.895 --> 00:20:40.337
One of the OG hardware hackers made a video.
315
00:20:40.377 --> 00:20:46.881
God, it had to have been like eight years ago at this point called like I hacked this crypto wallet and recovered three million dollars.
316
00:20:46.921 --> 00:21:00.471
Basically, it was a video showcasing him performing a software glitching fault injection attack on a Trezor wallet to help a gentleman retrieve three million dollars that he locked up and forgot the passkey for the wallet.
317
00:21:00.511 --> 00:21:03.733
OK, so we did it, got the money out, and I'm sure he got some kind of commission out of that.
318
00:21:03.833 --> 00:21:04.714
Good for you, Joe Grant.
319
00:21:05.334 --> 00:21:18.242
So if that takes that amount of effort to recover the key or change the software in the Trezor wallet, which was subject to that video, how is it possible that hackers are uploading malicious firmware to these wallets, right?
320
00:21:18.262 --> 00:21:19.443
Which is what's going on in this attack.
321
00:21:20.003 --> 00:21:26.667
Well, what's crazy is they're actually just completely de-soldering the
microchip on the Trezor wallet that is locked.
322
00:21:26.747 --> 00:21:31.150
The JTAG fuse has been blown on that thing, and they just put a new one down.
323
00:21:31.770 --> 00:21:41.325
And because the software for the Trezor wallet is open source, they compile a new Trezor wallet firmware with just compromised seed phrases.
324
00:21:41.385 --> 00:21:47.034
So the seed phrase is predictable so that they can offline reproduce that wallet and sign transactions from it.
325
00:21:47.943 --> 00:22:03.216
insane absolutely crazy when i was reading this i was like this isn't possible it's not possible that they're able to flash new code to a treasure wallet oh no it literally is if you replace the microcontroller completely which is what they've done so kudos to them i guess but also like jesus christ insane
326
00:22:03.516 --> 00:22:14.123
Well, the broader part of this here, right, is that a ton of crypto wallet, soft wallet, hard wallet, a ton of crypto companies jumped all over this Ledger story.
327
00:22:14.544 --> 00:22:16.705
And they were like, come to us.
328
00:22:16.785 --> 00:22:17.406
Ledger sucks.
329
00:22:17.466 --> 00:22:17.886
Look at that.
330
00:22:17.906 --> 00:22:19.627
Like, we're completely immune to this.
331
00:22:19.687 --> 00:22:21.588
We're like, we've never had this happen.
332
00:22:21.709 --> 00:22:21.909
Right.
333
00:22:22.189 --> 00:22:23.250
Trezor was one of them.
334
00:22:23.890 --> 00:22:36.619
And like, uh, MG literally says in this thread, like conveniently ignoring Joe grand's research where he found implanted treasure wallets in the wild, not even like theoretically in his research.
335
00:22:36.819 --> 00:22:37.519
I wasn't aware of that.
336
00:22:37.720 --> 00:22:38.260
Okay.
337
00:22:38.620 --> 00:22:42.443
So like found in the wild treasure implanted wallets as well.
338
00:22:42.703 --> 00:22:48.068
So, yeah, it's like, no, you're not immune to this.
339
00:22:48.088 --> 00:22:51.792
A bunch of these other crypto companies.
340
00:22:53.333 --> 00:22:59.699
He even says, then you have other competitors shilling their product while having an unspoken list of other problems.
341
00:23:00.620 --> 00:23:02.142
The crypto space is so sketchy.
342
00:23:02.222 --> 00:23:02.742
I agree, MG.
343
00:23:02.762 --> 00:23:02.942
I agree.
344
00:23:03.002 --> 00:23:03.323
I agree.
345
00:23:04.083 --> 00:23:04.863
It is super weird.
346
00:23:05.103 --> 00:23:08.344
Such a weird wild west unregulated environment.
347
00:23:08.844 --> 00:23:14.585
You know, not that regulations solve everything or don't cause their own problems, but it's just wild here, man.
348
00:23:14.625 --> 00:23:18.266
Especially when people are doing this with like their life savings.
349
00:23:18.946 --> 00:23:19.226
Yeah.
350
00:23:19.646 --> 00:23:24.187
You have so many people that are not treating this as like a hobby or like a side store of cash.
351
00:23:24.227 --> 00:23:28.728
You have so many people that, you know, lose their crypto and like, that's it.
352
00:23:28.788 --> 00:23:29.388
They're ruined.
353
00:23:29.448 --> 00:23:30.989
And it's like, oh man, this is rough.
354
00:23:31.109 --> 00:23:31.929
This is rough out here.
355
00:23:32.409 --> 00:23:37.451
Yeah, and so I don't really know the terminology too well.
356
00:23:37.631 --> 00:23:46.635
I think a potential solution is like an NFC cold wallet where it's like the key is not really even in an electronic device.
357
00:23:46.675 --> 00:23:53.018
It's truly just like the stored crypto key, but you have to like manually enter it somewhere else.
358
00:23:53.038 --> 00:23:58.480
I'm not sure if this solves a problem, but the issue here is that you have an electrical device that contains your private key.
359
00:23:58.600 --> 00:24:17.306
right and like it's going to get hacked that's that's kind of just the whole moral of the story here right is there there is a hack for everything um so you know maybe there aren't solutions to this either way it's uh super interesting for sure yeah and the treasure it's kind of wild the treasure implants that joe grand found in the wild was just in august wasn't even that long ago oh wow
360
00:24:19.415 --> 00:24:22.339
Yeah, the issue here, I guess, is that, like, the chip that they're using is so common.
361
00:24:22.359 --> 00:24:23.801
It's an STM32F.
362
00:24:24.281 --> 00:24:32.292
You can go buy those online, and if you have a heat gun, you can just go desolder and resolder, and then suddenly, you know, you've implanted the Trezor wallet.
363
00:24:32.312 --> 00:24:36.097
Obviously, you have to build the firmware and stuff, but, I mean, not super complicated.
364
00:24:36.537 --> 00:24:40.158
Seems like the similarity is a untrusted third-party reseller.
365
00:24:40.558 --> 00:24:40.698
Yep.
366
00:24:41.538 --> 00:24:41.878
Buy direct.
367
00:24:42.218 --> 00:24:43.219
Yeah, buy direct on these.
368
00:24:43.879 --> 00:24:45.159
All right, crypto.
369
00:24:45.599 --> 00:24:47.960
There's a few weeks in a row.
370
00:24:48.140 --> 00:24:52.481
For two non-crypto bros, we wind up covering a lot of these stories.
371
00:24:52.741 --> 00:24:55.701
I may not be crypto-pilled, but you know what pilled I am?
372
00:24:56.102 --> 00:24:57.682
Maze, the sponsor of today's video.
373
00:24:58.497 --> 00:24:59.037
Take it away.
374
00:24:59.318 --> 00:25:02.860
Hey, hackers, before we keep going, I want to take a quick break and say a quick thank you to our sponsors.
375
00:25:03.140 --> 00:25:06.663
Today's Lowdown episode is sponsored by our friends over at Maze.
376
00:25:07.143 --> 00:25:12.447
Maze has spent the last couple of years using AI to find and remediate vulnerabilities in the cloud.
377
00:25:12.547 --> 00:25:19.292
And now they've released Maze Code, AI agents that find and deeply investigate vulnerabilities in your code.
378
00:25:19.532 --> 00:25:20.333
Now, we all know the deal.
379
00:25:20.733 --> 00:25:26.982
Software composition analysis tools
and SaaS toolings are all historically really bad at just making more noise.
380
00:25:27.022 --> 00:25:33.893
They find issues that may not actually be real and just create a lot of problems for your engineers to deal with that don't solve real security vulnerabilities.
381
00:25:33.973 --> 00:25:34.393
I get it.
382
00:25:34.433 --> 00:25:34.994
We've been there.
383
00:25:35.635 --> 00:25:41.017
Maze code is a really cool tool because Maze code actually knows how your code works.
384
00:25:41.358 --> 00:25:46.420
Maze code investigates every finding with the context of how your code is actually deployed.
385
00:25:46.820 --> 00:25:54.423
Sure, you're using a version of libxml2 that's vulnerable, but the actual code path to hit that vulnerability is not exposed in your Docker image.
386
00:25:54.683 --> 00:25:58.185
Maze code can find that and won't show you that that vulnerability is a big deal.
387
00:25:58.385 --> 00:26:02.948
Every vulnerability that Maze code reports comes with evidence that it's real and a way to fix it.
388
00:26:03.288 --> 00:26:07.651
Check out Maze at go.thelowdownpod.com slash maze.
389
00:26:07.751 --> 00:26:11.454
That's go.thelowdownpod.com slash maze.
390
00:26:11.614 --> 00:26:12.234
Thanks again, Maze.
391
00:26:12.355 --> 00:26:15.196
Let's keep going.
392
00:26:15.216 --> 00:26:15.857
And we're back.
393
00:26:15.917 --> 00:26:19.239
You know, that Maze integration was pretty cool.
394
00:26:19.800 --> 00:26:23.142
You know what else is supposed to be cool, Matt, but wasn't?
395
00:26:24.367 --> 00:26:46.941
vanilla ice refrigerators ah refrigerators matt they're supposed to be cool but in france where they got all those baguettes and croissants i'm told uh not not a lot of not a lot of cold food over there what happened man i feel like we just need like i need a picture of someone just standing in front of a broken fridge smoking a cigarette it's like yeah i don't know
396
00:26:48.462 --> 00:26:49.943
Ah, jacobus.
397
00:26:50.043 --> 00:26:50.464
I don't know.
398
00:26:50.484 --> 00:26:51.064
Jacobus, yeah.
399
00:26:51.104 --> 00:26:52.825
I don't know why we're dumping on the French.
400
00:26:52.885 --> 00:26:53.366
It's kind of fun.
401
00:26:53.386 --> 00:26:53.966
I don't know.
402
00:26:54.206 --> 00:26:54.987
It is kind of fun.
403
00:26:55.007 --> 00:26:57.389
Sorry for our massive French audience.
404
00:26:57.429 --> 00:26:58.610
We're so American.
405
00:26:58.650 --> 00:26:59.490
It's crazy.
406
00:26:59.650 --> 00:27:00.691
We're really big in France.
407
00:27:01.311 --> 00:27:02.712
So yeah, I made a short about this.
408
00:27:02.732 --> 00:27:04.173
I had to talk about this.
409
00:27:04.193 --> 00:27:05.414
This got me a little ruffled.
410
00:27:05.434 --> 00:27:06.395
Why are you ruffled, Matt?
411
00:27:06.415 --> 00:27:07.136
Ask me why I'm ruffled.
412
00:27:07.176 --> 00:27:07.796
Why are you fluffled?
413
00:27:07.816 --> 00:27:08.617
Why are you fluffled?
414
00:27:09.386 --> 00:27:14.732
The normie media, once again, talking about cybersecurity stuff.
415
00:27:15.033 --> 00:27:18.837
Guess what they claimed was going on here?
416
00:27:19.278 --> 00:27:20.239
I saw your shorts.
417
00:27:20.279 --> 00:27:24.885
I feel like I have insider information, but could it potentially be that an AI superhacker did it?
418
00:27:25.549 --> 00:27:26.751
The AI super hacker.
419
00:27:26.931 --> 00:27:27.492
AI super hacker.
420
00:27:27.512 --> 00:27:27.892
Classic.
421
00:27:27.912 --> 00:27:31.938
So obviously people are like pointing at, whoa, a hack.
422
00:27:32.259 --> 00:27:38.828
And I know one thing that's going on right now is AIs are escaping the lab and hacking everything.
423
00:27:38.848 --> 00:27:42.113
Do you know what the actual vulnerability here is, everybody?
424
00:27:42.765 --> 00:27:43.525
You're not going to believe it.
425
00:27:43.785 --> 00:27:44.525
You're not going to believe it.
426
00:27:45.226 --> 00:27:50.407
A management portal open on the internet with a really stupid vulnerability.
427
00:27:50.707 --> 00:27:51.707
Hooray!
428
00:27:51.767 --> 00:27:54.088
The year of 2026 and we're still doing this.
429
00:27:55.528 --> 00:27:56.128
Oh, gosh.
430
00:27:56.228 --> 00:28:01.969
You know, I guess we could consider the grocery and food supply chain is also critical infrastructure.
431
00:28:02.010 --> 00:28:02.730
So here we are.
432
00:28:03.050 --> 00:28:06.250
We're a critical infrastructure management interface on the internet.
433
00:28:06.711 --> 00:28:09.271
And the vulnerability is as dumb as they get.
434
00:28:09.431 --> 00:28:11.432
This isn't even like, oh, a CVE.
435
00:28:13.132 --> 00:28:17.794
The username is admin and the password automatically changes every day.
436
00:28:18.454 --> 00:28:19.474
Really secure, right?
437
00:28:19.634 --> 00:28:19.854
Yes.
438
00:28:19.874 --> 00:28:21.475
Because we're going to rotate our password.
439
00:28:21.515 --> 00:28:22.955
That actually follows NIST guidance.
440
00:28:22.975 --> 00:28:23.515
That's perfect.
441
00:28:23.695 --> 00:28:24.376
Oh my gosh.
442
00:28:25.136 --> 00:28:25.976
To today's date.
443
00:28:26.156 --> 00:28:27.297
The password is today's date.
444
00:28:27.637 --> 00:28:27.977
Damn it.
445
00:28:28.957 --> 00:28:31.118
I'm so close.
446
00:28:31.258 --> 00:28:38.980
But did you know that today's date, if put into Unix timestamp format, uses numbers and special characters and letters, uppercase and lowercase.
447
00:28:39.040 --> 00:28:41.081
Technically, they're following NIST guidance.
448
00:28:41.462 --> 00:28:54.745
I had someone, because I went on my soapbox about this, and I had someone in my comments say, well, no, it is actually AI-related, Matt, because now everyone wants to spend money on AI, so they're not spending money on security.
449
00:28:55.385 --> 00:29:00.306
Brother, we haven't been spending money on security since before you were a glimmer in your parents' eye.
450
00:29:00.827 -->
00:29:02.947
We have been fighting for budget forever.
451
00:29:02.967 --> 00:29:04.447
No, this has nothing to do with AI.
452
00:29:04.507 --> 00:29:05.288
This is just dumb.
453
00:29:05.308 --> 00:29:06.268
Yeah.
454
00:29:06.808 --> 00:29:10.349
I will say there is potentially an AI edge to this in that...
455
00:29:10.709 --> 00:29:11.129
I hate you.
456
00:29:12.830 --> 00:29:21.355
Taking apart the software to find this vulnerability, if not previously disclosed, is easier, is now commoditized.
457
00:29:21.375 --> 00:29:23.217
This is someone that I know recently gave a keynote for.
458
00:29:24.597 --> 00:29:27.639
But what's crazy about this is I saw – yeah, I know, drink.
459
00:29:27.980 --> 00:29:30.962
I saw this pop off one or two days ago at this point.
460
00:29:31.702 --> 00:29:43.104
And I was scrolling LinkedIn, and then I saw Clarity, Team 82, they do a lot of great intel on, or research on, you know, bugs in routers, bugs in OT equipment, operational technology.
461
00:29:43.465 --> 00:29:51.466
They wrote an article on, wow, you know, being able to hack and defrost food in Dan Frost refrigeration control units.
462
00:29:52.246 --> 00:29:56.287
Surely this article came out within the last week or so related to this incident.
463
00:29:56.327 --> 00:29:58.788
No, brother, this was a four, no, two-month-old.
464
00:29:59.408 --> 00:30:24.159
research paper two months ago they were like hey by the way these bugs exist and on like props everybody in in that like uh in that article because they reported it to dan frost i'm assuming i'm pronouncing that correctly uh and they actually patched the firmware and shipped a release so that bug went away congratulations but also in critical infrastructure fashion we didn't update and we left it facing the internet insane yes matt
465
00:30:25.164 --> 00:30:25.404
Yeah.
466
00:30:25.584 --> 00:30:26.004
Oh, what?
467
00:30:26.205 --> 00:30:27.765
And we left it and we didn't patch it?
468
00:30:28.045 --> 00:30:28.446
Oh, man.
469
00:30:28.466 --> 00:30:31.167
This is completely different than everything that's ever happened.
470
00:30:31.407 --> 00:30:31.627
Yeah.
471
00:30:31.927 --> 00:30:32.327
Oh, man.
472
00:30:32.447 --> 00:30:35.429
It's just the same story, man, in a different shape.
473
00:30:35.789 --> 00:30:36.029
Okay.
474
00:30:36.209 --> 00:30:50.536
In like seemingly related incident, like unrelated in this exact vulnerability, but for some reason on the same day, a giant like umbrella corp that's over a bunch of different grocery chains was...
475
00:30:51.196 --> 00:30:58.138
had to take a bunch of their back office systems offline because of a cyber attack that we have not a lot of details about yet.
476
00:30:58.558 --> 00:31:15.744
But what are the chances that we've got like a ton of refrigerators at grocery stores going offline due to really dumb cyber attack at the same time as a different chain of grocery suppliers is going off at the same time?
477
00:31:16.772 --> 00:31:20.613
Not a lot of details on this one, but really it smells like ransomware, right?
478
00:31:20.633 --> 00:31:36.397
I mean, I don't want to like claim that too, obviously, but when you see these big retail companies with a very specific subset of their IT systems that are all of a sudden unavailable and then like their point of sale is fine.
479
00:31:37.097 --> 00:31:39.499
but they're unable to, like, order food.
480
00:31:40.580 --> 00:31:45.443
And so it's like, oh, their fresh food is going to suffer for the next couple days.
481
00:31:46.624 --> 00:31:48.245
I don't know what the hell's going on.
482
00:31:48.605 --> 00:31:49.606
Yeah, it's interesting.
483
00:31:49.846 --> 00:31:59.593
I saw a Twitter thread related to this, kind of, like, talking about the before, during, and after era of, like, technology.
484
00:31:59.633 --> 00:32:05.878
And it's like, okay, before this, we had to all farm our own food and grow our own food because, like, we didn't have the technology to support this kind of system, right?
485
00:32:05.898 --> 00:32:06.058
Yeah.
486
00:32:06.458 --> 00:32:12.899
And then we were able to with computers and, you know, computers helped automate all of this through farming and delivery and logistics.
487
00:32:12.939 --> 00:32:13.639
And they put it in the store.
488
00:32:13.659 --> 00:32:15.360
Now you can buy it with e-commerce, stuff like that.
489
00:32:15.820 --> 00:32:23.781
And then the question is like, OK, with with AI coming online and people being able to like just find vulnerabilities and everything and like act and everyone can act like an APT.
490
00:32:24.321 --> 00:32:31.863
Is there a world where like we kind of revert back to the old times where like out of a fear or necessity, you know, everyone's got to have their own farm again.
491
00:32:32.103 --> 00:32:32.583
You know what I mean?
492
00:32:32.603 --> 00:32:33.223
Like, where does this go?
493
00:32:33.543 --> 00:32:34.003
Kind of scary.
494
00:32:35.602 --> 00:32:41.885
Um, I literally late breaking news, new, new, uh, info on this right now.
495
00:32:41.905 --> 00:32:44.767
Uh, some other researchers are going in and looking.
496
00:32:44.867 --> 00:32:46.808
It's not just Dan Frost anymore.
497
00:32:46.828 --> 00:32:47.888
Uh, there's some other stuff.
498
00:32:47.908 --> 00:32:48.849
Uh, there's some other stuff.
499
00:32:50.831 --> 00:32:52.632
Copeland and some other vendors.
500
00:32:53.352 --> 00:33:06.480
And in research now, because we know about these vulnerabilities being exploded in the wild, we have other internet exposed controllers from another vendor that are already compromised by a botnet and have...
501
00:33:07.440 --> 00:33:11.223
Crypto miners already installed on them, but in the same kind of attack.
502
00:33:11.263 --> 00:33:19.469
So, you know, these same devices, instead of just knocking it to 50 degrees and trolling a whole country's food supply.
503
00:33:19.869 --> 00:33:20.750
Trolling
is not the right word.
504
00:33:20.790 --> 00:33:21.370
I'm so sorry.
505
00:33:22.070 --> 00:33:25.353
But yeah, just hacking in that regard.
506
00:33:25.833 --> 00:33:28.135
Now, all of a sudden, it's crypto miner botnet.
507
00:33:29.916 --> 00:33:31.657
Wow, I'll put this link in here too.
508
00:33:32.357 --> 00:33:43.362
I do find it kind of funny that like, it seems like no matter what is compromised and like the level of effect they could have on the planet, there's always one actor that's like, and crypto.
509
00:33:43.382 --> 00:33:43.862
Like, really?
510
00:33:43.922 --> 00:33:45.543
You could like do a lot of damage with that.
511
00:33:45.563 --> 00:33:47.424
You're just gonna mine Ethereum on this box?
512
00:33:47.844 --> 00:33:48.124
Really?
513
00:33:48.164 --> 00:33:48.945
Is that all you're gonna do?
514
00:33:48.985 --> 00:33:49.585
It's kind of crazy.
515
00:33:50.045 --> 00:33:51.906
Yeah, it happens all the time.
516
00:33:52.066 --> 00:33:52.927
Is that just North Korea?
517
00:33:53.007 --> 00:33:55.029
Is that just them trying to fund their nation?
518
00:33:55.049 --> 00:33:55.909
Like, how does that work?
519
00:33:55.929 --> 00:34:00.572
No, because I even think they, miners are too inefficient for them.
520
00:34:00.592 --> 00:34:01.573
They're stealing billions.
521
00:34:01.613 --> 00:34:04.415
What are you getting from a miner on a botnet for a little while, right?
522
00:34:04.515 --> 00:34:06.557
It's like a rounding error for them at this point.
523
00:34:06.797 --> 00:34:07.177
That's true.
524
00:34:08.998 --> 00:34:09.278
Yeah, man.
525
00:34:09.299 --> 00:34:11.260
So, I mean, I guess we keep talking a lot about like,
526
00:34:13.559 --> 00:34:19.222
companies, organizations, states, whatever, leaving infrastructure online, like on the internet.
527
00:34:19.922 --> 00:34:21.843
One, why does this keep happening?
528
00:34:22.284 --> 00:34:24.185
And two, how do we make it not happen?
529
00:34:24.705 --> 00:34:26.646
Like, why is this a repeated theme?
530
00:34:26.666 --> 00:34:27.767
And it's not like a new thing.
531
00:34:27.787 --> 00:34:29.528
This has been going on since like the dawn of time.
532
00:34:29.548 --> 00:34:31.048
You would just think that it would end by 2026.
533
00:34:32.887 --> 00:34:33.287
Oh man.
534
00:34:33.447 --> 00:34:43.374
So there's also a Reddit thread of people, staff working at these grocery stores in the US that are saying that they are hearing and obviously take this with a grain of salt.
535
00:34:43.434 --> 00:34:59.624
What is, you know, what is actual grocery staff is probably ransomware recovery is probably not there in their scope of work, but they're getting told it's four weeks out from recovery, which puts them right in the middle of Thanksgiving.
536
00:34:59.644 --> 00:34:59.744
Yeah.
537
00:35:00.244 --> 00:35:00.845
Oh, no.
538
00:35:01.005 --> 00:35:03.126
And that they're not going to have inventory.
539
00:35:03.326 --> 00:35:10.171
So let's keep an eye on this one to see if a ransomware attack disrupts grocery store shelving inventory during Thanksgiving.
540
00:35:11.131 --> 00:35:12.372
Would get some attention, wouldn't it?
541
00:35:12.412 --> 00:35:13.913
That's how you get some legislation, though.
542
00:35:13.933 --> 00:35:15.814
You know, Mitch McConnell can't have his pumpkin pie.
543
00:35:15.854 --> 00:35:17.595
We finally get something done about this.
544
00:35:17.635 --> 00:35:18.236
Maybe that'd be cool.
545
00:35:18.496 --> 00:35:19.997
Can they liquefy pumpkin pie?
546
00:35:24.595 --> 00:35:25.995
Not a political podcast, Matt.
547
00:35:26.075 --> 00:35:26.575
Stop it.
548
00:35:26.836 --> 00:35:27.136
Sorry.
549
00:35:27.156 --> 00:35:27.556
Put it out.
550
00:35:27.576 --> 00:35:33.937
Yeah, we don't have to talk about politics to make fun of Gerber-eating legislators.
551
00:35:34.137 --> 00:35:34.437
You know?
552
00:35:34.557 --> 00:35:35.177
Like, come on.
553
00:35:35.698 --> 00:35:36.498
Jesus Christ.
554
00:35:36.598 --> 00:35:37.138
Hey, what?
555
00:35:37.398 --> 00:35:37.898
There's another...
556
00:35:38.378 --> 00:35:42.919
Isn't the other party the guy just announced for re-election he's going to be 102 years old if he gets elected?
557
00:35:42.939 --> 00:35:44.379
I mean, it's just... Is that real?
558
00:35:44.580 --> 00:35:44.960
Yeah.
559
00:35:45.240 --> 00:35:45.820
That's crazy.
560
00:35:45.940 --> 00:35:46.700
It's dumb, man.
561
00:35:46.860 --> 00:35:47.580
He just announced...
562
00:35:47.880 --> 00:35:48.700
I forget his name.
563
00:35:48.820 --> 00:35:49.121
Whatever.
564
00:35:49.261 --> 00:35:50.381
One of the other old guys.
565
00:35:51.434 --> 00:35:52.176
Too much, too much.
566
00:35:52.256 --> 00:35:53.118
Okay.
567
00:35:53.780 --> 00:35:57.368
You know what else 102-year-olds are really bad at?
568
00:35:57.929 --> 00:35:58.711
What's that, Matt?
569
00:35:58.731 --> 00:35:59.773
You prefer such a deer.
570
00:36:04.038 --> 00:36:05.439
That's so stupid.
571
00:36:05.619 --> 00:36:06.039
Why?
572
00:36:06.140 --> 00:36:07.140
Our transitions.
573
00:36:07.460 --> 00:36:07.941
Oh, man.
574
00:36:08.061 --> 00:36:08.501
I'm so sorry.
575
00:36:09.161 --> 00:36:09.601
Oh, no.
576
00:36:09.741 --> 00:36:16.145
Well, actually, actually, actually, in the era of AI, take a shot every time that I say that.
577
00:36:16.185 --> 00:36:17.426
They actually could be pretty good at it.
578
00:36:17.546 --> 00:36:18.166
Okay, okay.
579
00:36:18.306 --> 00:36:21.528
I want to give you – we've gone so far off the rails today.
580
00:36:21.708 --> 00:36:24.370
I want to give you just a ton of credit.
581
00:36:24.531 --> 00:36:25.792
I know I kind of gave you some kudos.
582
00:36:26.272 --> 00:36:35.540
You did this balance that you and I have to strike as like practitioners in cybersecurity and technology for a very long time now, but also content creators.
583
00:36:36.000 --> 00:36:39.043
And like, we want to talk like really technical
stuff.
584
00:36:39.624 --> 00:36:47.791
And we have to strike this balance of being able to explain this to people that might not have ever done reverse engineering or might not have ever actually even looked at
585
00:36:48.271 --> 00:36:50.432
like a dump of a memory dump or something like that.
586
00:36:50.452 --> 00:36:52.834
But then we have to explain it on the internet, right?
587
00:36:52.874 --> 00:36:53.654
And like get it right.
588
00:36:54.114 --> 00:36:56.296
And a lot of times I do this in 90 seconds, right?
589
00:36:56.356 --> 00:36:58.857
And so I get a lot of comments that are like, that wasn't quite accurate.
590
00:36:58.917 --> 00:37:00.598
I'm like, you do better in 90 seconds.
591
00:37:00.618 --> 00:37:02.279
Yeah, if it's so easy, then you do it, please.
592
00:37:02.339 --> 00:37:02.999
I swear to God.
593
00:37:03.139 --> 00:37:04.600
And so like, whatever, right?
594
00:37:04.800 --> 00:37:06.081
But I got to give you kudos.
595
00:37:06.181 --> 00:37:08.963
We want to do a lot of the technical stuff, right?
596
00:37:09.023 --> 00:37:15.566
And sometimes we kind of have to like zoom in and out of that because maybe the most technical stuff doesn't perform very well on these platforms.
597
00:37:15.746 --> 00:37:16.367
You seem to have...
598
00:37:17.107 --> 00:37:20.909
somehow struck gold with both of your goals.
599
00:37:20.969 --> 00:37:25.670
And you made a really technical video that's doing really, really well right now on reverse engineering.
600
00:37:25.710 --> 00:37:28.852
I watched the whole thing to actually pull up those memory dumps and stuff like that.
601
00:37:28.912 --> 00:37:29.512
Super cool.
602
00:37:29.812 --> 00:37:31.273
Just kudos, kudos, kudos.
603
00:37:31.493 --> 00:37:32.033
I appreciate that.
604
00:37:32.213 --> 00:37:38.715
Yeah, what's cool about all this AI stuff, whether you like it or don't like it, is that for me, make this about me.
605
00:37:39.775 --> 00:37:50.577
Reverse engineering is like now in vogue, which is kind of cool, which means that like the average person, the normies, if you will, are interested in like, okay, how does software work?
606
00:37:50.677 --> 00:37:53.698
How do we take it apart and how can we create a new thing?
607
00:37:54.078 --> 00:38:01.900
And what's really great about this is it's all originating from a deep, justified, guttural hatred for the Adobe Corporation.
608
00:38:02.480 --> 00:38:17.045
Now, what we're seeing is that there's a bunch of new copies of software like the OGs, Photoshop, Premiere Pro, Lightroom popping up that are not one-for-one copies.
609
00:38:17.065 --> 00:38:23.227
There are a lot of performance issues with them, but they have a lot of the same features, and they are completely vibe-coded.
610
00:38:23.667 --> 00:38:28.391
Now, what's happening is there's this new project that is popping off on GitHub.
611
00:38:28.431 --> 00:38:29.873
It's called Reverse Engineer Anything.
612
00:38:29.913 --> 00:38:40.402
It's
literally just an MCP server that uses traditional reverse engineering tooling like Ida, Ghidra, and Hopper to decompile a project.
613
00:38:41.243 --> 00:38:43.924
and then recompile it from that pseudocode.
614
00:38:43.965 --> 00:38:45.145
And it does a really good job.
615
00:38:45.585 --> 00:38:58.732
I'll pull a picture of it up here on the podcast, but for the video that I'm working on that goes out tomorrow or the day after, I had it just kind of like do this process on GNOME calculator, kind of like the default calc that pops up there that shifts with Ubuntu.
616
00:38:59.032 --> 00:39:01.654
And obviously, low level, that's open source, bad idea.
617
00:39:01.734 --> 00:39:05.276
Shut up, I know, it was from the compiled version, not the decompiled version, or not from the source code.
618
00:39:06.016 --> 00:39:10.660
And it created in 28 minutes a one-for-one copy of a 986K elf.
619
00:39:10.700 --> 00:39:13.382
Like, it's really, really good.
620
00:39:14.283 --> 00:39:22.350
Now, this does, in my opinion, beg some sort of existential questions about software as a whole and also security, right?
621
00:39:22.650 --> 00:39:26.573
The big one for this podcast is, like, we talked about supply chain and, like, the Trezor.
622
00:39:27.494 --> 00:39:30.375
uh, and ledger scenarios that applies here, right?
623
00:39:30.395 --> 00:39:33.477
Like what does Adobe have on the line?
624
00:39:33.877 --> 00:39:43.642
If they start to either ship malware intentionally or get compromised and unwittingly unknowingly ship malware, well, then I don't pay them anymore.
625
00:39:43.682 --> 00:39:45.543
And they, as a corporation cease to exist.
626
00:39:45.563 --> 00:39:45.783
Right.
627
00:39:46.143 --> 00:39:50.065
But if Joe Schmo who vibe coded Photoshop too, right.
628
00:39:50.365 --> 00:39:50.425
If,
629
00:39:50.625 --> 00:39:54.430
If they are pushing malware intentionally or don't, there's no downside for them.
630
00:39:54.470 --> 00:40:00.718
So the like the model of trust with software is kind of turning where it's like everyone vibe coding everything and sharing it to everyone.
631
00:40:01.079 --> 00:40:05.805
There's no like root of trust anymore that I can lean on to be like, oh, well, surely Jimmy wouldn't send me malware.
632
00:40:05.825 --> 00:40:06.646
Well, no, he totally couldn't.
633
00:40:06.666 --> 00:40:07.707
It wouldn't affect him at all.
634
00:40:08.448 --> 00:40:11.769
And then also another issue is maintainability, right?
635
00:40:11.830 --> 00:40:24.335
Like, who's going to fix the inevitable, I don't know, XML parsing bug in Premiere Pro 2 Vibe Coded if they don't have profit behind it to fund the bug fixes, right?
636
00:40:25.015 --> 00:40:30.359
This has been a problem for, you know, coming up on 30 years or so in the open source world too, right?
637
00:40:30.399 --> 00:40:32.861
Who was incentivized to fix open source software?
638
00:40:32.901 --> 00:40:35.803
Well, it's literally just a couple of guys who
do it for the love of the game.
639
00:40:36.483 --> 00:40:38.965
And that's not like super sustainable in my opinion.
640
00:40:38.985 --> 00:40:40.346
So, yeah, man, what are your thoughts on this?
641
00:40:41.527 --> 00:40:44.269
Yeah, when I think of trust...
642
00:40:45.432 --> 00:40:51.116
I think of Adobe because they have not shipped me any bad code in the history of time.
643
00:40:51.316 --> 00:40:54.838
And they would never charge me $100 for canceling my subscription early.
644
00:40:54.878 --> 00:40:56.359
That'd be crazy.
645
00:40:56.599 --> 00:40:57.460
I do get your point.
646
00:40:57.540 --> 00:41:05.065
It's just funny to put Adobe up on the trust pedestal due to them being a whatever.
647
00:41:05.125 --> 00:41:07.066
Anyway, yeah, this one's interesting.
648
00:41:07.967 --> 00:41:11.789
I'd say there's like one YouTube channel that I feel...
649
00:41:12.730 --> 00:41:16.252
the most scared by existing because they're so good.
650
00:41:16.752 --> 00:41:18.333
And it's FireShip.
651
00:41:18.533 --> 00:41:19.213
Do you know FireShip?
652
00:41:19.593 --> 00:41:20.113
I love FireShip.
653
00:41:20.594 --> 00:41:21.494
Yeah, I do too.
654
00:41:21.734 --> 00:41:29.198
And that's why I'm scared of them existing because I love them because they're faceless and they could, you know, replicate.
655
00:41:29.638 --> 00:41:33.860
Like that channel could turn into a fully AI-controlled channel tomorrow if AI was good enough.
656
00:41:33.880 --> 00:41:34.340
Just to be clear.
657
00:41:34.700 --> 00:41:38.443
That is a rumor that's already happened and he has addressed it and it is not true.
658
00:41:38.704 --> 00:41:46.230
But every time he puts a video out, they're like, I'll wait for the Fireship video in the Fireship comments making a joke that it's not actually Jeff anymore.
659
00:41:46.250 --> 00:41:47.191
It's like an AI bot.
660
00:41:47.491 --> 00:41:48.492
Okay, I didn't know that.
661
00:41:48.852 --> 00:41:49.373
I didn't know that.
662
00:41:49.573 --> 00:41:50.994
And I love his channel.
663
00:41:52.535 --> 00:41:58.277
And, but it makes me nervous that it's so good because it's like, who's, you said, Jeff, I didn't know Jeff.
664
00:41:58.637 --> 00:41:59.598
I didn't know that his name.
665
00:41:59.798 --> 00:42:00.958
I don't, I don't know this person.
666
00:42:01.038 --> 00:42:01.859
I have no trust.
667
00:42:02.299 --> 00:42:08.761
Like the, the, the thing that you and I are hinging a lot of our time and energy on is that we're building some trust with some audience.
668
00:42:08.901 --> 00:42:10.662
Yes, correct.
669
00:42:10.882 --> 00:42:13.763
You can argue that's really all my business has is trust.
670
00:42:14.123 --> 00:42:18.986
It's like that they know us and that they know that we've done this for a career and like they're building this relationship with us.
671
00:42:19.106 --> 00:42:19.907
Okay.
672
00:42:20.087 --> 00:42:27.412
So he did a video on this and I got to watch him try to use the free... What is it?
673
00:42:27.472 --> 00:42:27.912
What's the...
674
00:42:28.232 --> 00:42:29.033
It's not Final Cut.
675
00:42:29.073 --> 00:42:29.313
What's...
676
00:42:29.913 --> 00:42:32.154
Oh, Film Cut or Film Now or something.
677
00:42:32.354 --> 00:42:33.635
No.
678
00:42:34.115 --> 00:42:34.996
What's the Adobe version?
679
00:42:35.016 --> 00:42:35.656
Oh, Premiere Pro.
680
00:42:35.796 --> 00:42:36.336
Premiere Pro.
681
00:42:36.476 --> 00:42:37.957
Sorry.
682
00:42:38.037 --> 00:42:39.618
It's called Filmcraft.
683
00:42:39.758 --> 00:42:40.898
So he used Filmcraft.
684
00:42:40.978 --> 00:42:46.341
He used the free reverse engineered Premiere Pro to try to do one of his videos.
685
00:42:46.361 --> 00:42:47.762
And that was going to be his whole thing.
686
00:42:47.782 --> 00:42:49.062
It's like, hey, I put out a video.
687
00:42:49.082 --> 00:42:51.103
I'm using the free one, right?
688
00:42:51.804 --> 00:42:52.944
He couldn't do it.
689
00:42:53.204 --> 00:42:54.185
Like, it didn't work.
690
00:42:54.485 --> 00:42:55.885
Like, so we're not there yet.
691
00:42:56.126 --> 00:42:56.366
Just FYI.
692
00:42:58.028 --> 00:43:00.513
But things are getting scarily possible.
693
00:43:00.713 --> 00:43:03.478
And especially when you're not trying to put out a...
694
00:43:04.769 --> 00:43:09.713
one-to-one, like, I'm going to make money on this or like, you know, corporate version of it.
695
00:43:10.193 --> 00:43:22.062
But as you see in security research, it's good enough already to like be able to do true security research of things that were previously closed source packaged appliances type thing, right?
696
00:43:22.262 --> 00:43:23.923
Which is scary in its own right.
697
00:43:24.424 --> 00:43:27.526
Actually, you know, I love pulling up my crystal ball, right?
698
00:43:27.546 --> 00:43:27.786
Yeah.
699
00:43:28.907 --> 00:43:31.269
that we were ahead of the curve on this.
700
00:43:31.549 --> 00:43:32.269
Well, not we.
701
00:43:32.309 --> 00:43:34.931
We shouldn't take credit, but we were on a panel at PlanetScale.
702
00:43:35.012 --> 00:43:35.192
Me.
703
00:43:35.712 --> 00:43:36.172
Yeah.
704
00:43:36.553 --> 00:43:37.674
We were at the Royal We.
705
00:43:38.374 --> 00:43:40.135
I was in the room when this was said.
706
00:43:40.236 --> 00:43:43.138
So I, therefore, transitive property.
707
00:43:44.439 --> 00:43:46.741
We did that panel at PlanetScale.
708
00:43:46.821 --> 00:43:47.481
When the hell was that?
709
00:43:47.501 --> 00:43:48.262
A couple months ago.
710
00:43:48.662 --> 00:43:48.822
Yeah.
711
00:43:49.142 --> 00:43:50.003
Time is flat circle.
712
00:43:50.543 --> 00:43:56.947
And that guy, Gabe from Sentinel one, we were, we were all chatting like before we even went on stage, uh, about stuff.
713
00:43:56.967 --> 00:44:09.795
And he was like, Hey, I think the actual risk of all of this is that like, it's going to get way too good at reverse engineering and the economy is going to explode because there's
no software moat anymore.
714
00:44:10.575 --> 00:44:11.776
Uh, and I was like, Oh man.
715
00:44:11.916 --> 00:44:12.216
All right.
716
00:44:12.356 --> 00:44:13.077
Like, that's interesting.
717
00:44:13.097 --> 00:44:14.998
I hadn't really thought about it a lot at that point.
718
00:44:15.358 --> 00:44:16.619
And now here we are.
719
00:44:16.739 --> 00:44:18.100
And like the cracks are starting to show.
720
00:44:18.120 --> 00:44:18.220
Yeah.
721
00:44:18.440 --> 00:44:19.340
Well, so if you...
722
00:44:19.421 --> 00:44:23.743
Okay, so again, I'm not trying to plug my own channel on this, but I am working on a video where I talk about this.
723
00:44:24.363 --> 00:44:29.166
My thoughts on that particular edge of the conversation are, yes, what is it?
724
00:44:29.186 --> 00:44:32.507
Like 70% of the S&P is software companies, which is not great.
725
00:44:32.968 --> 00:44:34.828
I have a lot of money in the S&P.
726
00:44:35.129 --> 00:44:36.549
Not a lot of money, but a big percentage.
727
00:44:36.589 --> 00:44:36.990
Fuck you guys.
728
00:44:37.310 --> 00:44:37.590
Anyway...
729
00:44:37.890 --> 00:44:39.391
Not to brag about how much money I have.
730
00:44:39.651 --> 00:44:41.771
I don't know, but a lot of my portfolio is in the S&P.
731
00:44:41.791 --> 00:44:46.193
What I'm trying to say is that I don't think corporations are going to stop being corporations.
732
00:44:46.213 --> 00:44:48.114
And I think they're always going to be profit incentivized.
733
00:44:48.154 --> 00:44:58.717
And so I think what will happen long term, and we're already seeing this happen in the games industry, where the user is just not going to get the software anymore.
734
00:44:59.298 --> 00:45:02.158
Look at games, PS5, a lot of that.
735
00:45:02.179 --> 00:45:05.100
What's happening is you don't get Grand Theft Auto 6.
736
00:45:05.240 --> 00:45:06.200
I'm making up an example.
737
00:45:06.620 --> 00:45:14.007
You get a ticket to Grand Theft Auto 6, you plug it into your PS5, and then you get, like, a cloud view of the game, right?
738
00:45:14.027 --> 00:45:19.271
You get literally just, like, a VDI, and the VDI lets you do, like, low-latency gaming, and that's it.
739
00:45:19.391 --> 00:45:22.494
Because they don't want you to have the media, because the media is kind of, like, their whole trade secret.
740
00:45:22.514 --> 00:45:27.658
And if you are a GTA 6, you've dumped, probably at this point, billions of dollars to lit it on fire.
741
00:45:27.678 --> 00:45:27.778
Okay.
742
00:45:27.818 --> 00:45:27.939
Okay.
743
00:45:28.599 --> 00:45:41.771
Now, I'm very nervous that software, the average consumer software, not even like games like Microsoft Word and like Photoshop and like this average shit that everyone uses will go the same path because Microsoft is going to need to make money still.
744
00:45:41.831 --> 00:45:43.012
Adobe is going to need to make money still.
745
00:45:43.192 --> 00:45:53.762
And so instead of giving you, you know, Photoshop.exe, they're just going to give you Photoshop remote thin client.exe and you plug into like Adobe servers and you get a view of it.
746
00:45:53.862 --> 00:45:57.070
Because otherwise, this is going to be the case for every piece of software.
747
00:45:57.251 --> 00:46:01.141
There are no trade secrets anymore if you give the user the trade secret.
748
00:46:01.800 --> 00:46:04.562
Yeah, I mean, the counterpoint to a lot of this, right?
749
00:46:04.582 --> 00:46:11.428
To like kind of not be doomerism about all of our money evaporating because software is not a moat anymore.
750
00:46:11.669 --> 00:46:12.349
Yeah.
751
00:46:12.790 --> 00:46:17.594
LibreOffice has existed for as long as I've, you know, been around.
752
00:46:17.694 --> 00:46:25.641
So it's not like Microsoft Word was Microsoft Word because it's the best at typing words into a square, right?
753
00:46:25.661 --> 00:46:25.881
Right.
754
00:46:26.201 --> 00:46:26.422
Like,
755
00:46:26.942 --> 00:46:29.383
The functionality can kind of exist.
756
00:46:29.403 --> 00:46:30.783
And we're already seeing it with this Adobe thing.
757
00:46:30.823 --> 00:46:33.584
It's like, yeah, you can get like 80% of the way there or whatever, right?
758
00:46:34.004 --> 00:46:46.387
The things that people wind up paying for is like the rest of the support and experience of being a customer of a company that's like paid to deliver this outcome to you.
759
00:46:46.407 --> 00:46:56.509
The outcome of using the Office Suite is, I've got a business that like purrs like a machine and like they're not going to stub their toe on some like random thing, right?
760
00:46:57.649 --> 00:46:59.890
that's what people are paying for.
761
00:47:00.070 --> 00:47:04.351
When a lot of the AI stuff started to take off, people were like, well, Salesforce is done, right?
762
00:47:04.411 --> 00:47:06.151
Like, I'm going to just code my own CRM.
763
00:47:06.191 --> 00:47:07.492
And it's like, people started to do it.
764
00:47:07.552 --> 00:47:11.553
And then you're like, oh, I now realize why I'm paying a company.
765
00:47:11.833 --> 00:47:12.953
Yeah, right.
766
00:47:13.333 --> 00:47:14.353
Even in our world, right?
767
00:47:14.413 --> 00:47:15.994
Like, we work with a bunch of editors.
768
00:47:16.074 --> 00:47:17.054
Those editors are...
769
00:47:17.834 --> 00:47:25.897
creating themselves more and more AI editor-y tool help to like speed up their job, right?
770
00:47:26.097 --> 00:47:28.878
I'm not paying them any less because I'm paying for an outcome.
771
00:47:29.058 --> 00:47:31.019
I'm paying for a good video on the other side of it.
772
00:47:31.079 --> 00:47:36.341
I don't care if you all of a sudden have a tool that does this in an eighth of the time.
773
00:47:36.601 --> 00:47:39.542
Like I was paying you for a quality output, right?
774
00:47:39.702 --> 00:47:40.643
And if your tool...
775
00:47:41.531 --> 00:47:49.041
Does, you know, 70% of a good job as you did, which is I think what
we're going to see here for a while with all of this, right?
776
00:47:49.562 --> 00:47:56.411
I'm still paying you for that 30% that like your experienced human touch is going to be able to cross the finish line with, right?
777
00:47:56.431 --> 00:47:56.932
Mm-hmm.
778
00:47:57.052 --> 00:48:00.174
So I think for a while, that's going to be the reality, right?
779
00:48:00.214 --> 00:48:07.777
Whether or not there's a ceiling on how good all of these models get, I've been wrong a few times before already, right?
780
00:48:08.218 --> 00:48:14.641
That the capability curve is going to start flattening and harnesses and human experience are still going to be a huge part of it just with a better tool.
781
00:48:15.041 --> 00:48:17.562
And then they've blown the roof off of capability.
782
00:48:17.602 --> 00:48:18.863
So who knows, right?
783
00:48:19.203 --> 00:48:20.704
But I still think there's like...
784
00:48:21.264 --> 00:48:24.406
hey, I'm paying a company and a person to handle this for me.
785
00:48:24.526 --> 00:48:25.826
It's like what I'm paying for, right?
786
00:48:26.307 --> 00:48:26.727
Yeah.
787
00:48:26.947 --> 00:48:27.787
So I don't know.
788
00:48:27.928 --> 00:48:40.694
But to your point, I'm seeing a lot of people come out and say, hey, in defense of SaaS now, this is now a pro for having a web app be your main interface for your software, right?
789
00:48:40.774 --> 00:48:41.755
Instead of shipping stuff out.
790
00:48:42.447 --> 00:48:42.627
Yeah.
791
00:48:43.228 --> 00:48:53.119
Yeah, I think this will actually probably more negatively hurt the gigantic companies and really, really help the midsize companies where there like truly is a person helping you out.
792
00:48:53.139 --> 00:49:00.567
Because like to counter your counter, like if you get help from Microsoft, Microsoft Word, you're not talking to a person for like eight layers of support.
793
00:49:00.627 --> 00:49:00.807
Right.
794
00:49:00.867 --> 00:49:01.348
It's just like.
795
00:49:01.508 --> 00:49:02.769
So not you, right?
796
00:49:02.949 --> 00:49:08.654
But I worked at a very large U.S. bank who was a very large Microsoft customer.
797
00:49:08.914 --> 00:49:10.055
I see.
798
00:49:10.255 --> 00:49:11.396
Do you know what I mean?
799
00:49:11.416 --> 00:49:12.157
Yeah, I got you.
800
00:49:12.737 --> 00:49:18.862
Do you think the bank was going to be like, actually, for our 330,000 employees, we're going to install LibreOffice?
801
00:49:19.082 --> 00:49:20.582
Sure.
802
00:49:20.663 --> 00:49:21.803
And like, that's it.
803
00:49:22.003 --> 00:49:22.203
Right.
804
00:49:22.323 --> 00:49:24.024
It's like a different scale of things.
805
00:49:24.164 --> 00:49:24.404
Right.
806
00:49:24.784 --> 00:49:28.245
And like, it's like, well, no, you know, I need patches.
807
00:49:28.285 --> 00:49:29.005
I need support.
808
00:49:29.605 --> 00:49:37.888
I can't have a team of people that make sure that LibreOffice is up to date and supported and new features and
security bugs and all this kind of stuff are happening.
809
00:49:38.228 --> 00:49:39.349
Then now that's my headache.
810
00:49:39.529 --> 00:49:39.749
Right.
811
00:49:40.089 --> 00:49:40.289
Yeah.
812
00:49:40.989 --> 00:49:42.210
This has always been the case, right?
813
00:49:42.350 --> 00:49:46.191
Open source is free like a puppy, not free like beer.
814
00:49:47.752 --> 00:49:48.452
You got to take care.
815
00:49:48.472 --> 00:49:49.233
Here's a free puppy.
816
00:49:49.433 --> 00:49:49.813
Oh, I see.
817
00:49:49.833 --> 00:49:50.873
I see.
818
00:49:52.054 --> 00:49:54.455
That's not the end of that transaction, right?
819
00:49:54.695 --> 00:49:56.896
Unlike free like beer, end of transaction.
820
00:49:57.496 --> 00:49:58.496
Everyone loves a free beer.
821
00:50:00.237 --> 00:50:00.517
Trust me.
822
00:50:01.137 --> 00:50:01.777
Yeah, yeah, yeah.
823
00:50:02.038 --> 00:50:02.838
You've never heard that saying?
824
00:50:03.178 --> 00:50:04.038
No, I like that, though.
825
00:50:04.618 --> 00:50:10.559
So does user commodity software go away and then just, like, corporate B2B stuff stays around?
826
00:50:10.599 --> 00:50:14.120
Are you saying that contract exists everywhere?
827
00:50:14.320 --> 00:50:15.620
I don't think anything goes away, right?
828
00:50:15.720 --> 00:50:22.161
I think the furthest right, like, oh, this is going to explode everything, is just never going to be true.
829
00:50:22.201 --> 00:50:30.403
Whether or not growth, like, also, infinite growth forever is not a really good model to base an entire economy on either, but here we are, right?
830
00:50:30.423 --> 00:50:31.083
Sure, sure, sure.
831
00:50:31.103 --> 00:50:31.363
So...
832
00:50:32.403 --> 00:50:35.545
So no, I mean, I don't know about going away or not.
833
00:50:35.705 --> 00:50:40.448
I just know like even watching some of this stuff with Adobe, like stuff go on.
834
00:50:40.688 --> 00:50:40.888
Right.
835
00:50:41.269 --> 00:50:49.213
And even in my personal life, I've like tried to replace some things that I use like with free alternatives.
836
00:50:49.834 --> 00:50:51.515
And the experience is so much more.
837
00:50:52.363 --> 00:50:57.648
we're like harder to get the same outcome that I'm like, all right, I'm going to throw 60 bucks at this.
838
00:50:57.708 --> 00:51:03.273
Like, especially like as more money is involved, like just in your life.
839
00:51:03.593 --> 00:51:06.456
So I'm not even talking about like, oh, as you make more money, right?
840
00:51:06.796 --> 00:51:13.622
Because for some people, the 60 bucks or whatever it is for a certain piece of software, that's a significant amount of money for a lot of people.
841
00:51:13.762 --> 00:51:14.583
That's not right.
842
00:51:14.663 --> 00:51:17.666
It's like, okay, like I spent more at Chili's this Friday, right?
843
00:51:17.926 --> 00:51:18.146
Right.
844
00:51:18.406 --> 00:51:21.612
Um, but for corporations, it's really not right.
845
00:51:21.772 --> 00:51:24.256
So like I run a business here, you run a business here.
846
00:51:24.517 --> 00:51:28.183
Like there's tons of stuff that I like, do I need to pay for Slack the way that I do?
847
00:51:28.363 --> 00:51:30.186
There's free chat alternatives.
848
00:51:30.247 --> 00:51:30.407
Right.
849
00:51:30.427 --> 00:51:30.828
But you do.
850
00:51:30.888 --> 00:51:31.689
And it's just great.
851
00:51:31.949 --> 00:51:32.150
Right.
852
00:51:32.795 --> 00:51:35.998
Like, you know, there's just these little things where I'm like, well, just pay for it.
853
00:51:36.018 --> 00:51:36.919
I don't want to think about this.
854
00:51:37.099 --> 00:51:41.283
Like, you know, my brain power is better used somewhere else, right?
855
00:51:41.303 --> 00:51:43.785
Yeah, what is the opportunity cost of your time to care about that?
856
00:51:43.805 --> 00:51:44.726
A hundred percent.
857
00:51:44.746 --> 00:51:45.867
Give someone else the problem.
858
00:51:46.087 --> 00:51:52.633
That's amplified so much when you're talking about corporations, when they're like, well, I would have to staff a team.
859
00:51:52.913 --> 00:51:53.974
Yeah, sure.
860
00:51:53.994 --> 00:51:56.877
To put this in security context, I'm sorry, I'm like up on a big soapbox now.
861
00:51:56.917 --> 00:52:02.381
To put it in a security context, I joined a pretty big security team that was on the Elk stack for a while, right?
862
00:52:02.582 --> 00:52:07.626
So free SIM, Elastic, you know, Kibana, like all this kind of stuff.
863
00:52:07.646 --> 00:52:16.234
I had like talented security engineers on my team that were spending like a significant percentage of their time being sysadmins, keeping the Elk stack standing.
864
00:52:16.294 --> 00:52:16.974
It would chip over.
865
00:52:17.034 --> 00:52:17.435
It would like...
866
00:52:17.835 --> 00:52:18.795
Just things would happen.
867
00:52:18.855 --> 00:52:21.356
You have to have sysadmins like managing the stuff.
868
00:52:21.616 --> 00:52:27.377
Eventually we were like, I want my security engineers doing security, not sysadmining my sim, right?
869
00:52:27.837 --> 00:52:30.378
So it's like, oh, this is why people pay Splunk.
870
00:52:30.578 --> 00:52:31.998
Everyone laughs about how much Splunk pays.
871
00:52:32.038 --> 00:52:39.820
And it's like, well, this turns out it's a really hard problem because you're throwing petabytes of logs at something that you have to index and keep alive.
872
00:52:39.900 --> 00:52:41.501
And if it goes down, everything breaks.
873
00:52:41.621 --> 00:52:43.161
And it's like, oh, okay, got it.
874
00:52:43.221 --> 00:52:44.721
This is why you write a check, right?
875
00:52:44.961 --> 00:52:45.762
Yeah.
876
00:52:45.862 --> 00:52:47.582
This is the same kind of thing, right?
877
00:52:47.962 --> 00:52:48.122
Yeah.
878
00:52:48.543 --> 00:52:48.923
Yeah, yeah, yeah.
879
00:52:49.303 --> 00:52:50.124
All right, Soapbox.
880
00:52:50.905 --> 00:52:57.890
No, first of all, for the record, Matt, this podcast is
mostly you and I getting on and off Soapboxes, so it's totally cool.
881
00:52:57.990 --> 00:52:58.470
No worries.
882
00:52:58.651 --> 00:52:59.131
That's true.
883
00:52:59.191 --> 00:52:59.591
That's true.
884
00:52:59.651 --> 00:53:00.712
That's like our job, right?
885
00:53:00.732 --> 00:53:01.973
Yeah, man.
886
00:53:02.033 --> 00:53:04.295
Well, sometimes I get a little...
887
00:53:04.795 --> 00:53:05.556
I have a few buttons.
888
00:53:05.596 --> 00:53:06.177
We found one.
889
00:53:06.357 --> 00:53:07.197
Sure, sure, sure.
890
00:53:07.618 --> 00:53:07.918
All right.
891
00:53:08.258 --> 00:53:09.339
So this one's crazy.
892
00:53:09.359 --> 00:53:10.200
Can I drive this one?
893
00:53:10.400 --> 00:53:10.700
Send it.
894
00:53:10.720 --> 00:53:11.341
Yeah, yeah, yeah.
895
00:53:11.361 --> 00:53:11.601
All right.
896
00:53:11.921 --> 00:53:13.582
This one's nuts because...
897
00:53:14.784 --> 00:53:16.248
Man, I have to make a short about this today.
898
00:53:16.268 --> 00:53:17.150
I was thinking about it.
899
00:53:17.170 --> 00:53:18.834
I was like, oh, I've been meaning to talk about this.
900
00:53:19.796 --> 00:53:21.239
I don't think I've...
901
00:53:22.283 --> 00:53:29.267
ever seen a less ethical creation of margin and profit margin in my life.
902
00:53:29.747 --> 00:53:31.208
Not even the Adobe Corporation?
903
00:53:31.668 --> 00:53:32.089
Whoa!
904
00:53:33.389 --> 00:53:34.410
That's called a callback.
905
00:53:34.650 --> 00:53:38.352
It's like a family guy tangent right there.
906
00:53:39.453 --> 00:53:40.934
So, okay.
907
00:53:41.534 --> 00:53:44.096
I also cannot believe that this happened twice in one week.
908
00:53:44.436 --> 00:53:45.857
So we have two...
909
00:53:46.637 --> 00:53:50.199
executives of ransomware negotiation firms.
910
00:53:50.840 --> 00:54:08.570
So this is like, if you're unaware, if you're a victim of ransomware and hackers have gotten in and encrypted all your sensitive stuff and you get that big, scary screen with skulls, that's like straight out of an NCS clip about hacking, that's like, uh, you're hacked.
911
00:54:09.090 --> 00:54:12.432
Here's my Bitcoin wallet in order to get your like data back.
912
00:54:13.202 --> 00:54:16.583
Well, people panic and they don't know who to call.
913
00:54:16.783 --> 00:54:17.963
You can't call Ghostbusters.
914
00:54:18.343 --> 00:54:27.346
So they call these ransomware negotiation services or recovery services or however they're marketed to be like, hey, I need some cybersecurity experts to come in and like solve this for me.
915
00:54:27.366 --> 00:54:27.486
Right.
916
00:54:27.506 --> 00:54:28.546
Because I don't know what the hell to do.
917
00:54:28.566 --> 00:54:29.326
I have a question.
918
00:54:29.566 --> 00:54:29.786
Yes.
919
00:54:29.846 --> 00:54:32.787
Matt, the guidance has changed much like children.
920
00:54:32.807 --> 00:54:34.927
Do the babies go on their belly or on their back?
921
00:54:34.967 --> 00:54:37.088
Do you pay the
ransom or do you not pay the ransom?
922
00:54:37.408 --> 00:54:47.371
This is a hotly contested topic, but the FBI's official stance, if they hold any credibility left anymore.
923
00:54:47.872 --> 00:54:49.312
Is that bestiality is okay.
924
00:54:49.352 --> 00:54:49.812
Got it.
925
00:54:49.852 --> 00:54:52.573
Besides that, dude, I missed that news cycle.
926
00:54:52.673 --> 00:54:56.134
And when I talked about the FBI getting hacked, my comments filled up with stuff like that.
927
00:54:56.154 --> 00:54:57.795
And I was like, I am missing something.
928
00:54:57.815 --> 00:54:59.856
I am very confused and scared.
929
00:54:59.976 --> 00:55:03.857
And I had to be very careful with what I Googled to figure out what those comments were about.
930
00:55:04.357 --> 00:55:06.117
so that I could tell bestiality.
931
00:55:06.598 --> 00:55:07.878
Ah, crap.
932
00:55:07.978 --> 00:55:09.518
I shouldn't have searched for this.
933
00:55:09.618 --> 00:55:09.818
Right.
934
00:55:10.259 --> 00:55:10.539
Okay.
935
00:55:10.679 --> 00:55:13.619
So, uh, yeah, FBI's guidance is to not pay.
936
00:55:13.979 --> 00:55:14.560
Right.
937
00:55:14.580 --> 00:55:17.440
For obvious reasons, you're a, you're funding a criminal enterprise.
938
00:55:17.580 --> 00:55:22.142
B, uh, these criminals are teenagers for a lot of the times.
939
00:55:22.722 --> 00:55:27.783
And if you give them millions of dollars, they tend to just be like, lol, I'm going to disappear.
940
00:55:27.823 --> 00:55:29.383
They disappear on their own crew.
941
00:55:29.423 --> 00:55:31.544
Sometimes they don't, they don't help.
942
00:55:31.964 --> 00:55:37.065
they don't care about helping the enterprise get back up if you just handed them like enough money to disappear.
943
00:55:37.545 --> 00:55:37.745
Right.
944
00:55:37.805 --> 00:55:39.746
So there's like a lot of reasons to not.
945
00:55:39.766 --> 00:55:41.826
There's no guarantee you'll actually get your data back.
946
00:55:42.006 --> 00:55:42.186
Sure.
947
00:55:42.626 --> 00:55:50.228
That being said, this was the whole shiny hunters thing was their PR campaign to be like, we're credible because they want their threats to be paid.
948
00:55:50.428 --> 00:55:54.469
So they need people to realize that they're actually not that bad.
949
00:55:54.669 --> 00:55:54.909
Right.
950
00:55:55.069 --> 00:55:57.810
And that they're not exaggerating in their claims, all this kind of stuff.
951
00:55:57.830 --> 00:55:59.050
So PR is important to them.
952
00:55:59.070 --> 00:55:59.310
Mm hmm.
953
00:56:00.477 --> 00:56:01.998
But official guidance is to not pay them.
954
00:56:02.278 --> 00:56:04.020
That being said, happens all the time.
955
00:56:04.080 --> 00:56:05.100
Happens every day, right?
956
00:56:05.541 --> 00:56:07.322
Because people are like, I don't know what the hell to do.
957
00:56:07.402 --> 00:56:10.504
You ransom a hospital without a security team.
958
00:56:11.325 --> 00:56:12.506
And they're like, I don't know.
959
00:56:12.746 --> 00:56:14.647
I need to get back online right now.
960
00:56:14.827 --> 00:56:17.309
Like, you know, what's the quickest path?
961
00:56:17.449 --> 00:56:22.072
Well, the quickest path is to send a bunch of cryptocurrency to that wallet address, right?
962
00:56:22.152 --> 00:56:22.753
Hope for the best.
963
00:56:23.742 --> 00:56:31.407
But what happened is there's this industry of intermediaries that popped up to help these companies that got hit with this ransomware, have no idea what to do.
964
00:56:31.927 --> 00:56:33.248
They go and Google, what do I do?
965
00:56:33.288 --> 00:56:35.650
And these companies pop up and they come in and try to help people.
966
00:56:36.210 --> 00:56:38.852
Well, there's a few of these companies and it turns out,
967
00:56:39.492 --> 00:56:46.534
that more than once now, some of these companies have been working both sides, both sides of the ransomware coin.
968
00:56:47.214 --> 00:56:51.395
Okay, so this Canadian guy living in Pennsylvania, the FBI picked him up.
969
00:56:51.735 --> 00:56:58.397
He was a ex-founder of Cypher, is I guess how you would say that, C-Y-P-C-H.
970
00:56:59.177 --> 00:56:59.638
F-E-R.
971
00:57:00.118 --> 00:57:00.338
Okay.
972
00:57:00.358 --> 00:57:02.920
But that's what his LinkedIn says is an ex-founder.
973
00:57:03.220 --> 00:57:08.605
Turns out that company, not too happy that he's out there calling himself a founder because they're like, he wasn't a founder.
974
00:57:08.985 --> 00:57:10.726
And they're trying to distance themselves from him.
975
00:57:10.746 --> 00:57:12.408
They're like, he was just a manager director here.
976
00:57:12.608 --> 00:57:13.008
Whatever.
977
00:57:13.188 --> 00:57:13.469
Right.
978
00:57:13.569 --> 00:57:18.112
He's also a published author of a book called Cyber Extortion Strategic Response.
979
00:57:19.974 --> 00:57:20.814
That's awesome.
980
00:57:22.736 --> 00:57:27.580
A practical manual for leaders on managing coercive events, negotiation, and settling strategies.
981
00:57:28.959 --> 00:57:29.419
Love that.
982
00:57:31.260 --> 00:57:33.980
I really want a copy of this book to put on my bookshelf here now.
983
00:57:34.520 --> 00:57:38.381
That's why I should fill this bookshelf with just books from cyber criminals.
984
00:57:38.601 --> 00:57:40.621
I'm sure there are plenty.
985
00:57:40.802 --> 00:57:44.782
He's gone out and spun up a different company now too.
986
00:57:44.802 --> 00:57:45.943
So not Cypher.
987
00:57:46.023 --> 00:57:48.923
Cybersteward is the other company.
988
00:57:49.703 --> 00:57:53.284
So Cypher, Cybersteward, they're kind of in the limelight on this.
989
00:57:54.184 --> 00:57:56.845
But he got picked up for...
990
00:57:57.745 --> 00:58:01.966
like by the FBI for being in on shiny hunters.
991
00:58:02.846 --> 00:58:05.647
So not just like, Oh, working both sides of the ransom coin.
992
00:58:05.827 --> 00:58:11.688
We've got this executive at a ransomware company that is all of a sudden in
cahoots with shiny hunters.
993
00:58:12.328 --> 00:58:17.729
If you're not a long time listener of the show, shiny hunters, we keep talking about, uh, just hacked the FBI.
994
00:58:17.749 --> 00:58:20.190
Uh, two of their leaders just got, uh,
995
00:58:20.830 --> 00:58:22.492
popped by the FBI as well.
996
00:58:22.612 --> 00:58:26.077
So it turns out, I think they're getting a lot of information on their investigation.
997
00:58:26.137 --> 00:58:29.882
Maybe people are flipping on each other is what I'm assuming is happening.
998
00:58:30.519 --> 00:58:34.960
we don't have a ton of detail about like what this guy did.
999
00:58:35.180 --> 00:58:36.481
Ed Dubrovsky.
1000
00:58:37.141 --> 00:58:41.182
We don't have a ton of information about what he was doing with shiny hunters.
1001
00:58:41.442 --> 00:58:42.503
Okay.
1002
00:58:42.563 --> 00:58:43.923
We have like some press releases.
1003
00:58:43.943 --> 00:58:44.623
We've got some stuff.
1004
00:58:44.643 --> 00:58:49.385
We just know that he's in custody and we have that he's in custody in relation to the shiny hunters thing.
1005
00:58:49.845 --> 00:58:54.567
So we'll let you know when we hear some like more info about like what this guy was doing.
1006
00:58:54.607 --> 00:58:55.427
Cybersecurity executive.
1007
00:58:55.647 --> 00:58:56.847
Dude, I covered this.
1008
00:58:56.927 --> 00:58:58.648
I had people DMing me that they worked with him.
1009
00:58:59.968 --> 00:59:05.931
They were like, I've run into this guy professionally, which is now the second shiny hunters person that I've caught those DMS about.
1010
00:59:05.951 --> 00:59:07.232
The last guy was a pen tester.
1011
00:59:07.252 --> 00:59:10.413
The guy that got picked up in the Netherlands or something, right?
1012
00:59:10.513 --> 00:59:10.733
Yeah.
1013
00:59:11.494 --> 00:59:12.914
I was a pen test consultant.
1014
00:59:13.035 --> 00:59:16.376
I had people being like, yo, I was a pen tester with that guy a few years ago.
1015
00:59:16.636 --> 00:59:17.076
Crazy.
1016
00:59:17.136 --> 00:59:18.257
This guy's a little older.
1017
00:59:20.418 --> 00:59:25.961
Shout out to VX Underground for what's the exact quote from VX Underground on this is...
1018
00:59:27.181 --> 00:59:28.882
Dang, I'm going to keep it 100 with you.
1019
00:59:28.962 --> 00:59:36.708
Never in a million years would I have expected a boomer maple syrup again to be an active member of extortion group named after a Pokemon.
1020
00:59:37.008 --> 00:59:37.549
Crazy.
1021
00:59:38.109 --> 00:59:48.957
Yeah, so obviously, you know, not truly a boomer, but significantly older than the average shiny hunters slash the calm threat actor.
1022
00:59:49.718 --> 00:59:52.340
Okay, we were prepping for the show.
1023
00:59:53.722 --> 00:59:56.463
And I was like, no, we have specifics about this.
1024
00:59:56.603 --> 01:00:01.825
I read specifics about this guy creating margin on these ransomware deals.
1025
01:00:02.245 --> 01:00:04.586
And I was like searching
all of the articles about this guy.
1026
01:00:04.606 --> 01:00:08.367
And I'm like, why are none of them talking about the specifics that I know that I read this week?
1027
01:00:09.968 --> 01:00:11.509
There was another one this week.
1028
01:00:12.910 --> 01:00:19.235
There was another executive of a ransomware helper group called Monster Cloud.
1029
01:00:20.196 --> 01:00:20.996
What's this guy's name?
1030
01:00:21.497 --> 01:00:21.957
I missed it.
1031
01:00:22.217 --> 01:00:23.858
Oh, Zohar Pinhazi.
1032
01:00:24.219 --> 01:00:24.539
Okay.
1033
01:00:24.879 --> 01:00:25.660
Also 50.
1034
01:00:25.760 --> 01:00:27.841
So also like older than your average thought actor.
1035
01:00:27.922 --> 01:00:29.483
Executive of Monster Cloud.
1036
01:00:29.603 --> 01:00:31.024
Owned and operated Monster Cloud.
1037
01:00:31.464 --> 01:00:41.767
which advertised themselves as a ransomware remediation company, that they said that they had tools and decryption techniques for recovering your encrypted data without paying cyber criminals.
1038
01:00:41.827 --> 01:00:54.871
Well, let's be very careful here, because Pinhazi very specifically told his employees not to use the word decryption and just proprietary remediation techniques, which include playing the middleman.
1039
01:00:54.891 --> 01:00:57.812
So it turns out all this guy was doing
1040
01:00:58.915 --> 01:01:03.359
was paying the threat actors for the decryption key and then charging the customers more.
1041
01:01:04.440 --> 01:01:05.121
How much more, Matt?
1042
01:01:05.642 --> 01:01:06.342
Oh, my God.
1043
01:01:06.402 --> 01:01:07.544
I had some specifics.
1044
01:01:07.864 --> 01:01:14.751
I'm pretty sure the number you told me pre-show was like, they paid the ransom group like $8,200 and charged the client like $100,000.
1045
01:01:14.791 --> 01:01:15.231
$150,000.
1046
01:01:15.251 --> 01:01:15.631
$150,000.
1047
01:01:15.651 --> 01:01:17.313
$150,000, which, to be fair, you've been in consulting, right?
1048
01:01:17.333 --> 01:01:17.914
Yeah, probably.
1049
01:01:22.779 --> 01:01:26.423
not an unheard of consulting SOW cost, right?
1050
01:01:26.463 --> 01:01:28.586
Yeah, but for actual work.
1051
01:01:28.846 --> 01:01:29.347
Right, right, right.
1052
01:01:29.387 --> 01:01:37.076
But I'm just saying, you can imagine a company gets a SOW from some sort of cybersecurity expert for...
1053
01:01:37.737 --> 01:01:42.121
you know, tens of thousands to low hundreds of thousands of dollars.
1054
01:01:42.161 --> 01:01:45.705
And they're like, yeah, I guess this is what cybersecurity costs, right?
1055
01:01:46.125 --> 01:01:51.170
Like that's what a pen, like a pen test can run you, you know, 60 to 150 grand for a pen test for a few weeks.
1056
01:01:51.230 --> 01:01:51.410
Right.
1057
01:01:51.730 --> 01:01:56.615
But the fact that all he was doing was paying the freaking cyber criminals a few thousand dollars, right?
1058
01:01:57.296 --> 01:02:01.940
Uh, in another one, he paid 236,000 and
charged the customer about 380,000.
1059
01:02:02.520 --> 01:02:10.948
So, uh, all in all, he charged clients more than $19 million and paid about 8 million in ransom payments.
1060
01:02:11.248 --> 01:02:11.508
All right.
1061
01:02:12.069 --> 01:02:13.970
So I'm going to get in trouble for asking this question.
1062
01:02:14.611 --> 01:02:17.574
Um, what did he do that is wrong?
1063
01:02:17.594 --> 01:02:17.654
Uh,
1064
01:02:19.585 --> 01:02:22.486
Because to me, no, because to me, it seems like he.
1065
01:02:22.867 --> 01:02:23.127
Okay.
1066
01:02:23.507 --> 01:02:23.967
Example.
1067
01:02:24.367 --> 01:02:27.249
My friend bought a Subaru Outback in 2018.
1068
01:02:27.269 --> 01:02:27.749
Okay.
1069
01:02:29.350 --> 01:02:32.832
But he didn't want to buy it himself.
1070
01:02:32.932 --> 01:02:36.273
He didn't want to negotiate with the car dealership because that's a fucking nightmare.
1071
01:02:36.614 --> 01:02:39.955
So he paid Costco to negotiate on his behalf.
1072
01:02:40.455 --> 01:02:41.596
And then Costco took a cut.
1073
01:02:42.256 --> 01:02:43.237
This is the same thing.
1074
01:02:44.664 --> 01:02:47.986
I think it's because he's advertising that he's not doing that, that it's fraud.
1075
01:02:48.366 --> 01:02:50.847
He didn't advertise that he had decryption.
1076
01:02:50.867 --> 01:02:54.649
He had proprietary remediation technology, which is not wrong.
1077
01:02:54.749 --> 01:02:55.549
Which is email.
1078
01:02:59.365 --> 01:03:11.128
No, so if you read the article, he was charged with one count of conspiracy to commit wire fraud and two counts of actual wire fraud in connection with an alleged ransomware decryption scheme.
1079
01:03:11.188 --> 01:03:16.430
So I'm assuming there's evidence somewhere that maybe he was like truly on the inside on the other side, maybe.
1080
01:03:16.450 --> 01:03:17.410
Yeah.
1081
01:03:17.610 --> 01:03:20.711
But without that, I really don't understand what's illegal here.
1082
01:03:20.971 --> 01:03:27.733
Honestly, the quote right there, as alleged in the indictment, by falsely claiming to decrypt ransomware without paying...
1083
01:03:29.073 --> 01:03:30.353
Without paying off the ransomware.
1084
01:03:30.373 --> 01:03:30.633
Yeah.
1085
01:03:30.713 --> 01:03:31.834
Sorry, let me say it again.
1086
01:03:32.094 --> 01:03:41.576
As alleged in the indictment by falsely claiming to decrypt ransomware without paying off the ransomware, the defendant re-victimized his clients while extracting a hefty profit for himself.
1087
01:03:41.917 --> 01:03:43.317
Re-victimized his clients.
1088
01:03:43.357 --> 01:03:44.037
Explain that.
1089
01:03:44.057 --> 01:03:51.399
Yeah, I guess he's saying basically that he's also now ransoming them, is what they're saying, right?
1090
01:03:51.419 --> 01:03:53.740
If I charge you to negotiate for me, you're not...
1091
01:03:54.860 --> 01:04:10.935
holding me at gunpoint this is a fee yeah yeah listen man I see your point it's obviously just like
he's obviously being a dick I'm just saying like if you zoomed out from this and like didn't call it ransomware and called it like home purchase what's the difference
1092
01:04:11.648 --> 01:04:15.932
Because I don't think that they're knowingly paying him to just go interface with this.
1093
01:04:15.972 --> 01:04:16.532
No, you're right.
1094
01:04:16.552 --> 01:04:16.813
You're right.
1095
01:04:16.833 --> 01:04:17.113
You're right.
1096
01:04:17.193 --> 01:04:19.335
They're paying him to come in, like get them recovered.
1097
01:04:19.415 --> 01:04:19.655
Right.
1098
01:04:19.915 --> 01:04:20.136
Yeah.
1099
01:04:20.556 --> 01:04:21.277
Oh my gosh.
1100
01:04:21.577 --> 01:04:22.318
Just so wild.
1101
01:04:22.618 --> 01:04:23.799
Yeah, it is crazy.
1102
01:04:23.819 --> 01:04:26.441
A big, big prison sentence lingering for him too.
1103
01:04:26.501 --> 01:04:27.502
It's like 20 years.
1104
01:04:27.522 --> 01:04:27.823
Yeah.
1105
01:04:27.903 --> 01:04:28.363
Yeah.
1106
01:04:28.563 --> 01:04:28.964
Yeah.
1107
01:04:29.864 --> 01:04:43.215
Anyway, it just seems like there's a pattern here that a lot of these ransomware firms are building closer relationships with the threat actors than they're leading off, including some of them working with them, right?
1108
01:04:43.455 --> 01:04:46.378
We have another very, very large insider story.
1109
01:04:46.798 --> 01:04:51.119
Yeah, seamless transition at a company you probably actually have heard of.
1110
01:04:51.299 --> 01:04:52.239
Wait, who is it?
1111
01:04:52.359 --> 01:04:53.239
Oh, OpenAI.
1112
01:04:53.639 --> 01:04:54.640
Oh, okay, OpenAI.
1113
01:04:54.680 --> 01:04:55.980
This is the funniest picture.
1114
01:04:56.000 --> 01:04:58.540
Yeah, this is the best.
1115
01:04:58.780 --> 01:05:00.701
So I didn't understand the context of this.
1116
01:05:00.821 --> 01:05:04.602
I thought this was a guy, like, detecting that he was getting spied on.
1117
01:05:05.242 --> 01:05:10.323
I did not realize that this picture was the spy committing the espionage
1118
01:05:11.043 --> 01:05:28.393
seeing that he was getting counter espionage getting tracked and it's him detecting the plant dude it's so funny he's like oh no i'm fucked just so many dumb criminal things here so of course by the way okay let's what what happened let's get to it before we get into the reaction so opening your eye
1119
01:05:28.953 --> 01:05:39.105
put out a statement that they've been cooperating with the FBI, that they detected that one of their employees that happens to be a Saudi national, planted a listening device in an executive conference room.
1120
01:05:39.866 --> 01:05:40.066
Okay.
1121
01:05:40.106 --> 01:05:41.608
They detected this a few ways.
1122
01:05:42.188 --> 01:05:48.375
One of which was that he left a bag at the office that had his wallet in it with his ID and the recording device.
1123
01:05:49.076 --> 01:05:53.178
So it's just like, okay, here's my ID next to my crime, right?
1124
01:05:53.298 -->
01:05:53.819
Very funny.
1125
01:05:54.399 --> 01:05:55.440
Then they went and found the device.
1126
01:05:55.760 --> 01:06:06.726
I guess it was a sting operation to him with this camera, like where he was going to go check on the device and got this really, really funny picture of him going to check on his recording device.
1127
01:06:07.846 --> 01:06:15.831
The other bit of evidence that they have is that from his wife's chat GPT account, couldn't even use Claude, couldn't even use Claude to do this,
1128
01:06:16.111 --> 01:06:16.471
Hold on.
1129
01:06:16.652 --> 01:06:17.632
I'm not tracking this part.
1130
01:06:17.973 --> 01:06:21.276
He used chat GPT to spy on open AI.
1131
01:06:21.356 --> 01:06:33.627
His wife's chat GPT account to ask about recording devices, like all of like the Google search stuff of like, how do I do this crime was on his wife's chat GPT account.
1132
01:06:33.767 --> 01:06:34.588
That's incredible.
1133
01:06:34.608 --> 01:06:35.469
I did not realize that.
1134
01:06:35.689 --> 01:06:42.995
According to asking the chat bot about building a continuous audio recorder, customizing Wi-Fi connectivity and encrypting the SD card.
1135
01:06:43.256 --> 01:06:56.707
And so I did notice this felt very off the shelf because the picture of the actual device uses a Seed Studio 2.4 gig antenna, which is like one of the, you know, resellers of like Makerspace Raspberry Pi esque things.
1136
01:06:56.807 --> 01:07:01.251
So like, yeah, this is literally a tutorial for baby's first listening device.
1137
01:07:01.371 --> 01:07:01.872
So, yeah.
1138
01:07:02.212 --> 01:07:06.956
So this isn't, you know, nation state capability level stuff here.
1139
01:07:07.136 --> 01:07:11.619
Obviously, the conspiracy theory people are coming out on this one, man.
1140
01:07:11.639 --> 01:07:14.041
I don't know if... Have you talked about this one publicly yet?
1141
01:07:14.101 --> 01:07:14.622
No, no, no.
1142
01:07:14.642 --> 01:07:15.743
I'm not tracking any conspiracies.
1143
01:07:16.143 --> 01:07:17.144
So I made some videos about it.
1144
01:07:17.164 --> 01:07:18.785
So they showed up in my comments, man.
1145
01:07:19.285 --> 01:07:20.766
They showed up.
1146
01:07:21.227 --> 01:07:24.749
I've been having to hit the translate button on my Twitter replies a lot.
1147
01:07:24.809 --> 01:07:29.153
So my tweet has made its way over in that direction, right?
1148
01:07:29.933 --> 01:07:32.734
Uh, and people are, are calling this, well, this was too easy.
1149
01:07:32.814 --> 01:07:33.134
Yeah.
1150
01:07:33.174 --> 01:07:34.454
Like he would ever be that dumb.
1151
01:07:34.474 --> 01:07:36.495
Uh, these evidence is all planted.
1152
01:07:36.515 --> 01:07:38.095
This is, you know, whatever.
1153
01:07:38.115 --> 01:07:42.616
And now, of course there's pictures of, of this guy with the crown prince of Saudi Arabia.
1154
01:07:42.776 --> 01:07:46.697
And so like the thread is starting to be pulled a lot on this one.
1155
01:07:47.137 -->
01:07:49.198
And people are like, yeah, right.
1156
01:07:49.358 --> 01:07:54.899
Like this is the worst, you know, fake false flag story I've ever heard.
1157
01:07:54.939 --> 01:07:57.199
And I'm like, Oh God, I can't, I can't possibly.
1158
01:07:59.480 --> 01:08:01.063
Dude, it's... No, there's no...
1159
01:08:01.083 --> 01:08:01.583
It's nonsense.
1160
01:08:01.623 --> 01:08:03.106
Okay, I'm like, who are they false flagging?
1161
01:08:03.126 --> 01:08:03.566
What do you mean?
1162
01:08:03.847 --> 01:08:05.109
Right, yeah, like, just...
1163
01:08:05.189 --> 01:08:05.549
I don't know.
1164
01:08:05.649 --> 01:08:08.353
We're just coming up with bad headlines about Saudi stuff.
1165
01:08:08.614 --> 01:08:14.102
It's also just, like, a damn shame that, like, tweets from, like, three weeks ago from OpenAI, like, corporate are, like...
1166
01:08:14.843 --> 01:08:17.985
profiling him as like their only Saudi employee.
1167
01:08:18.005 --> 01:08:24.150
And they're like, look at this, like, you know, putting him up on a pedestal and it's like, oh, fast forward.
1168
01:08:24.570 --> 01:08:25.771
You're probably wondering how I got here.
1169
01:08:27.112 --> 01:08:33.897
You know, like is in FBI custody, like literally three weeks after they were like pedestaling him for all this stuff.
1170
01:08:33.957 --> 01:08:35.579
So this one's gnarly, obviously.
1171
01:08:35.599 --> 01:08:42.345
You know, this is part of... We talked about that whole LG story with the TVs having microphones.
1172
01:08:42.385 --> 01:08:53.836
And one of the actual threats there, instead of advertisers listening to you in your living room, is the TVs with microphones in executive boardrooms that have remote code execution in them.
1173
01:08:54.397 --> 01:08:55.118
And like...
1174
01:08:55.778 --> 01:08:58.300
A threat actor could compromise and tap into that microphone.
1175
01:08:58.320 --> 01:08:59.480
Well, here you go, right?
1176
01:08:59.580 --> 01:09:06.624
Here's like someone actually trying to tap into the audio of an executive conference room via a TV just behind it, right?
1177
01:09:06.684 --> 01:09:07.285
Yep.
1178
01:09:07.345 --> 01:09:08.345
With an external device.
1179
01:09:08.666 --> 01:09:11.707
I mean, this is exactly what we were talking about.
1180
01:09:11.727 --> 01:09:12.648
I think the other...
1181
01:09:13.268 --> 01:09:14.889
Like picture here.
1182
01:09:15.190 --> 01:09:17.831
And we were talking about a little bit as we were getting ready.
1183
01:09:18.952 --> 01:09:20.253
I've run and you have some opinions.
1184
01:09:20.313 --> 01:09:21.274
I'm actually interested in.
1185
01:09:21.454 --> 01:09:29.700
I've run a lot of my like mental math on building security programs geared towards being able to catch insider threats.
1186
01:09:30.643 --> 01:09:36.028
Because if you are... Not that you don't trust the adults that you hire or anything like that, right?
1187
01:09:36.568 --> 01:09:42.974
I'm a firm believer in trusting the people that you hire and treating them like adults and letting them do stuff and not micromanaging and all this kind of stuff.
1188
01:09:43.615 --> 01:09:56.486
But when it comes to security of things, and when I come to monitoring behavior on my cloud, my network, my code, whatever it is, my mental model is what if...
1189
01:09:57.447 --> 01:10:00.870
insider, then now I'm watching this behavior.
1190
01:10:00.890 --> 01:10:04.173
So it's not just like, oh, that account is my super senior cloud guy.
1191
01:10:04.193 --> 01:10:06.174
So whatever he's doing is probably fine.
1192
01:10:06.474 --> 01:10:06.735
No.
1193
01:10:07.175 --> 01:10:07.415
Right.
1194
01:10:07.635 --> 01:10:08.576
What's the context?
1195
01:10:08.596 --> 01:10:09.337
What's going on?
1196
01:10:09.697 --> 01:10:12.119
Why is that account running those weird commands in prod?
1197
01:10:12.379 --> 01:10:13.340
All these kinds of things.
1198
01:10:13.480 --> 01:10:13.720
Right.
1199
01:10:14.101 --> 01:10:21.607
And so I think that it's kind of an unspoken thing in the industry until these stories pop up is that this stuff is pretty prevalent.
1200
01:10:22.067 --> 01:10:31.950
Like there are lots of spies working in Silicon Valley companies, you know, whether or not they all get caught in such a headline grabbing way or not is, is, is one thing.
1201
01:10:32.370 --> 01:10:46.314
But I've talked to CISOs out there and they just kind of have it in their head that like a certain percentage of people that work for these big tech companies, uh, might be sending whatever information they can get, uh, to, you know, another country.
1202
01:10:47.123 --> 01:10:48.004
Yeah, no, I'm with you.
1203
01:10:48.064 --> 01:10:49.825
I don't think we diverge at all on that.
1204
01:10:49.845 --> 01:10:50.565
I completely agree.
1205
01:10:50.585 --> 01:10:51.586
And I think we mentioned this before.
1206
01:10:51.626 --> 01:10:57.450
Like when you are on a private network that is owned by another corporation, you lose all of your constitutional rights, right?
1207
01:10:57.470 --> 01:10:58.870
Like you are getting spied on.
1208
01:10:58.930 --> 01:11:08.076
You cannot do the things you want to do because, you know, the company is trying to make sure that you're not going to, you know, implant the device of the CEO and spy on their communications.
1209
01:11:08.116 --> 01:11:08.617
I'm cool with that.
1210
01:11:09.362 --> 01:11:26.692
I think my reluctance to completely agree with like this attitude is that sometimes that posture, I think, bleeds into network policy and infrastructure policy that makes the execution of your job very painful.
1211
01:11:26.712 --> 01:11:33.376
And it damages employee morale and that you don't feel like you're trusted as an adult.
1212
01:11:33.716 --> 01:11:37.300
You're not able to do your job because the network is too tight because, oh, it's not you.
1213
01:11:37.740 --> 01:11:38.701
It's the insider.
1214
01:11:38.801 --> 01:11:41.444
And it's
like, okay, bro, but like fucking who am I?
1215
01:11:41.524 --> 01:11:41.984
You know what I mean?
1216
01:11:42.445 --> 01:11:43.186
So I'm with you.
1217
01:11:43.206 --> 01:11:44.747
I don't think they diverge that far.
1218
01:11:44.787 --> 01:11:50.693
I'm very sensitive about that mentality going too far in one direction of that.
1219
01:11:51.433 --> 01:11:57.635
Yeah, I've worked places that it's really hard to get your job done because it's like you're just... Oh, you want a pip install?
1220
01:11:57.775 --> 01:11:59.116
I don't know about that.
1221
01:11:59.136 --> 01:12:02.477
We talk a lot about supply chains like H.E.R.D., but mirrors exist?
1222
01:12:02.517 --> 01:12:03.577
I don't know.
1223
01:12:03.757 --> 01:12:05.217
It's complicated.
1224
01:12:05.598 --> 01:12:10.419
I was the head of security for a fintech company that was a small company here in Austin that got bought by Goldman Sachs.
1225
01:12:11.519 --> 01:12:19.362
It took us weeks to get back into prod post-acquisition because it was like every command we ran on like...
1226
01:12:19.982 --> 01:12:21.143
the Kubernetes command line.
1227
01:12:21.203 --> 01:12:24.565
It was like, oh, you run kubectl and like a bunch of stuff happens.
1228
01:12:24.925 --> 01:12:25.725
All of it was blocked.
1229
01:12:26.426 --> 01:12:27.947
And so it's like, oh, you hit a wall.
1230
01:12:27.967 --> 01:12:29.007
You had to like submit a ticket.
1231
01:12:29.047 --> 01:12:30.448
Be like, oh, can you unblock this thing?
1232
01:12:31.068 --> 01:12:32.029
No, we can't unblock this thing.
1233
01:12:32.049 --> 01:12:33.730
Google Cloud is obviously not allowed.
1234
01:12:34.090 --> 01:12:36.572
It's like, well, you just bought a company that's hosted on Google Cloud.
1235
01:12:36.612 --> 01:12:38.393
And like, can you escalate this?
1236
01:12:38.513 --> 01:12:39.773
Okay, yeah, escalated.
1237
01:12:39.974 --> 01:12:42.135
Forgot these guys in Austin have approval for that.
1238
01:12:42.595 --> 01:12:44.716
But then they would just approve that one URL.
1239
01:12:45.197 --> 01:12:47.318
And then you'd run the command again and you'd hit the next URL.
1240
01:12:47.358 --> 01:12:48.158
And it's like, no, I need this domain.
1241
01:12:48.178 --> 01:12:49.219
And then you're just playing for a game.
1242
01:12:49.679 --> 01:12:53.800
You're playing ticket Olympics to get the guy at the sock and the knock to approve your shit.
1243
01:12:53.840 --> 01:12:57.061
And then again, they're treating you adversarially, unintentionally.
1244
01:12:57.221 --> 01:12:59.581
It comes off as like you're the bad guy.
1245
01:12:59.621 --> 01:13:01.582
You have to justify everything.
1246
01:13:01.762 --> 01:13:04.903
And then it's like when a cop is following you and you've done nothing wrong.
1247
01:13:04.923 --> 01:13:06.823
And it's like, do I have cocaine in the car?
1248
01:13:06.943 --> 01:13:07.903
You know what I mean?
1249
01:13:08.143 --> 01:13:09.284
It just makes me feel like that.
1250
01:13:09.344 --> 01:13:09.744
It sucks.
1251
01:13:11.105 --> 01:13:12.526
I can't relate to that last one.
1252
01:13:12.546 --> 01:13:12.746
Sorry.
1253
01:13:12.766 --> 01:13:13.247
I don't know.
1254
01:13:13.267 --> 01:13:14.307
I'm just kidding.
1255
01:13:14.347 --> 01:13:14.928
But yeah.
1256
01:13:15.008 --> 01:13:15.808
I don't do cocaine.
1257
01:13:15.829 --> 01:13:17.630
It's an analogy.
1258
01:13:18.630 --> 01:13:20.031
But yeah, I get it, right?
1259
01:13:20.232 --> 01:13:27.076
But in terms of like when I'm building a security program and monitoring is like my detection is my goal.
1260
01:13:27.317 --> 01:13:28.237
Yeah.
1261
01:13:28.437 --> 01:13:33.281
Gearing towards insider threat as a detection mindset is really smart in my opinion.
1262
01:13:33.561 --> 01:13:34.201
No, I totally agree.
1263
01:13:34.241 --> 01:13:34.942
I have no issue with that.
1264
01:13:35.671 --> 01:13:36.671
All right, man, we went long.
1265
01:13:37.212 --> 01:13:40.153
There's one last story we could touch on really quick.
1266
01:13:40.573 --> 01:13:45.515
You and I are not iOS experts, but there's some drama in iOS jailbreak land.
1267
01:13:45.535 --> 01:13:48.196
So we wanted to bring it up because it seems like it's some crazy exploits.
1268
01:13:48.556 --> 01:13:50.097
We do not have the exploit details.
1269
01:13:50.437 --> 01:13:52.398
So we can't really make like a fun segment on it.
1270
01:13:53.218 --> 01:14:04.381
But it seems like people on iPhone 18 Pro Maxes on iOS 27.2 have been able to figure out some ways to install some side-loaded stuff on their devices.
1271
01:14:04.441 --> 01:14:05.781
But it seems like it's not reliable.
1272
01:14:06.221 --> 01:14:07.341
Maybe there's some...
1273
01:14:08.302 --> 01:14:10.542
They're not commercializing it is what they're getting at.
1274
01:14:10.662 --> 01:14:12.563
And so, like, we can't go look at it.
1275
01:14:13.123 --> 01:14:13.723
It exists.
1276
01:14:13.743 --> 01:14:14.423
Yeah.
1277
01:14:14.483 --> 01:14:16.984
Go read the research if you're into iOS stuff.
1278
01:14:17.724 --> 01:14:20.445
We'll keep an eye on it if we get any more info on these.
1279
01:14:20.786 --> 01:14:22.606
And, like, there's actual exploits.
1280
01:14:22.746 --> 01:14:25.608
The exploits seem to be server-side, though, which is really, really interesting.
1281
01:14:25.908 --> 01:14:26.228
Yeah.
1282
01:14:26.588 --> 01:14:40.454
The deal we are here is that, like, jailbreaks in iOS land, or really in any land, relies on a bug in the kernel or the hypervisor in, like, an enclave system that allows you to run code at the most privileged level so you can take over the rest of the device.
1283
01:14:41.054 --> 01:14:47.940
What's interesting about this jailbreak is that it's not actually a vulnerability in, like, the kernel of your device or a secure enclave of your device.
1284
01:14:48.421 --> 01:14:57.749
It's actually – it was a vulnerability in the way that Apple's developer
program does signing on code when you're doing development.
1285
01:14:57.830 --> 01:15:00.112
And something about it they didn't whitelist.
1286
01:15:00.632 --> 01:15:10.162
what things they would sign and allow them effectively to ship a firmware update that got them control over privileged devices within the iPhone.
1287
01:15:10.182 --> 01:15:19.250
I think like the, like the hardware enclave or like the iOS architecture is very segmented and has a lot of different trust layers, which is good.
1288
01:15:19.891 --> 01:15:27.959
But I can't speak intelligently about it because to understand this article, you have to be like a 10-year iOS insider, which I am not.
1289
01:15:28.199 --> 01:15:30.441
So anyway, if you want to read it, we'll put it in the link for the show.
1290
01:15:30.561 --> 01:15:31.462
Really interesting stuff.
1291
01:15:31.562 --> 01:15:32.343
Also very long.
1292
01:15:32.383 --> 01:15:35.046
So I'm trying my best to like boil it down, but doing a poor job.
1293
01:15:35.166 --> 01:15:38.369
But yeah, pretty neat research it looks like.
1294
01:15:38.609 --> 01:15:39.430
So there you go.
1295
01:15:39.865 --> 01:15:40.245
Yeah, man.
1296
01:15:40.706 --> 01:15:41.386
All right.
1297
01:15:41.726 --> 01:15:44.028
So, cool thing about our next episode.
1298
01:15:44.448 --> 01:15:44.949
Mm-hmm.
1299
01:15:45.189 --> 01:15:46.170
We get to do it in person.
1300
01:15:46.550 --> 01:15:46.870
It's true.
1301
01:15:47.150 --> 01:15:49.452
I get to, like... We will be... What?
1302
01:15:49.492 --> 01:15:49.912
Touch you.
1303
01:15:49.953 --> 01:15:50.893
I get to smell you.
1304
01:15:50.913 --> 01:15:52.594
I get to, like... No, I don't know.
1305
01:15:52.815 --> 01:15:53.075
Sorry.
1306
01:15:53.615 --> 01:15:54.316
That was weird.
1307
01:15:54.516 --> 01:15:55.236
That was weird, Matt.
1308
01:15:56.237 --> 01:15:57.858
Um, well, we'll be in Vegas to be at one con.
1309
01:15:57.978 --> 01:16:01.161
Uh, we have a coupon code to get the one con for a little cheaper.
1310
01:16:01.201 --> 01:16:02.102
I believe it's still active.
1311
01:16:02.122 --> 01:16:02.802
If it is, put it here.
1312
01:16:02.822 --> 01:16:04.483
And if not, Josh, we'll cut this out of the show.
1313
01:16:04.984 --> 01:16:14.731
Uh, but yeah, come hang out with us at one con and, uh, you know, we're gonna have some good interviews with some of the people over there at Sutton, the one, some of the presenters and talk about the future of security in the era of AI.
1314
01:16:14.791 --> 01:16:16.072
Take a shot every time I say that.
1315
01:16:16.632 --> 01:16:17.752
But in the meantime, guys, do us a favor.
1316
01:16:17.772 --> 01:16:20.153
If you're watching this, if you're listening to this, it really does help the show.
1317
01:16:20.193 --> 01:16:29.715
If you like, comment, subscribe on YouTube, if you rate us five stars on Spotify, we're actually getting discovered natively in the Spotify algorithm, according to some of my friends, which is pretty cool.
1318
01:16:29.735 --> 01:16:31.575
And that only happens if you rate the show.
1319
01:16:32.015 --> 01:16:37.336
And then also best way to help the show out in total is check out the sponsor of the show, Maze, good partner of the show.
1320
01:16:37.356 --> 01:16:42.037
They really are out there trying to make the internet a safer place with all these AI super hackers taking over.
1321
01:16:42.537 --> 01:16:44.418
So as usual, guys, thank you for listening.
1322
01:16:44.438 --> 01:16:45.058
We appreciate it.
1323
01:16:45.118 --> 01:16:45.898
And we'll see you.
1324
01:16:46.439 --> 01:16:46.881
In the next one.
1325
01:16:47.284 --> 01:16:47.646
Toodaloo.
1326
01:16:48.089 --> 01:16:48.270
Peace.