رونوشت 🔗
1
00:00:00,000 --> 00:00:05,180
All right, gentlemen, thank you so much for joining us. This was quite the momentous weekend.
2
00:00:05,620 --> 00:00:13,100
We had the largest supply chain attack in crypto history. And you two were sort of in the middle
3
00:00:13,100 --> 00:00:18,700
of it all, discovering it, documenting it, giving your takes and giving insights to the community,
4
00:00:19,180 --> 00:00:26,100
showing us what exactly happened. So I'd like to start with Mark. You are the former CEO of
5
00:00:26,100 --> 00:00:32,800
Mt. Gox, but this week you were the person to catch this giant fraud and attack happening.
6
00:00:33,420 --> 00:00:35,620
How did you end up in the center of all this?
7
00:00:36,020 --> 00:00:41,960
Well, it's quite a long story, but I have had people asking me about ledgers because
8
00:00:41,960 --> 00:00:44,540
they have coins locked into their ledgers.
9
00:00:44,660 --> 00:00:45,900
So I looked into it.
10
00:00:45,900 --> 00:00:50,540
I was trying to get a few ledgers on my own to try a few things.
11
00:00:50,540 --> 00:01:00,220
And I found very, very cheap ledgers on some marketplaces and also on Amazon Japan, which led me to think that this looks suspicious.
12
00:01:00,800 --> 00:01:05,540
So I placed the orders, actually placed three orders.
13
00:01:05,860 --> 00:01:06,840
One never came.
14
00:01:08,040 --> 00:01:11,360
One was a Chinese seller, but the ledger came from Malaysia.
15
00:01:12,660 --> 00:01:16,280
And one was a Japanese seller, but he actually bought it from Amazon.
16
00:01:16,460 --> 00:01:17,660
So I'm guessing it's the same seller.
17
00:01:17,660 --> 00:01:24,800
but what came in when it came in I actually immediately opened it and found that it was
18
00:01:24,800 --> 00:01:30,860
including something that was not part of the normal ledger so I've had a few other open
19
00:01:30,860 --> 00:01:37,640
ledgers at the time so I could very easily compare each one and find that it was very wrong
20
00:01:37,640 --> 00:01:41,720
yes so we will dive into the technicals and everything of how it all went down
21
00:01:41,720 --> 00:01:44,600
But I'd like to hand it over to Zach right now.
22
00:01:44,700 --> 00:01:48,960
I mean, you were last with us on this show for the cold card hack.
23
00:01:49,340 --> 00:01:53,400
And now here we are with Ledger facing a massive exploit.
24
00:01:54,180 --> 00:01:58,740
I mean, what went through your head when you first saw Mark's photos on X?
25
00:01:59,300 --> 00:02:03,580
Well, firstly, I just be clear, I've done nothing compared to Mark.
26
00:02:04,820 --> 00:02:09,400
I've just been involved in replying to tweets and talking about it.
27
00:02:09,400 --> 00:02:12,180
Mark, I can't imagine how exhausted you must be.
28
00:02:12,800 --> 00:02:13,960
It's good to meet you.
29
00:02:14,960 --> 00:02:17,880
I just think that in these couple months,
30
00:02:17,880 --> 00:02:21,780
all of these attacks that have been more theoretical,
31
00:02:22,640 --> 00:02:26,080
discussed by the hardcore Bitcoiners,
32
00:02:26,720 --> 00:02:31,400
retirement attacks, they called the entropy type bug,
33
00:02:31,820 --> 00:02:33,140
supply chain attacks,
34
00:02:33,700 --> 00:02:35,400
we're seeing all of this manifest
35
00:02:35,400 --> 00:02:37,600
in a very compressed period of time.
36
00:02:37,600 --> 00:02:40,120
and I think people are really scared
37
00:02:40,120 --> 00:02:43,300
and I think some of the reactions have been
38
00:02:43,300 --> 00:02:47,100
just like the reaction to the cold card incident
39
00:02:47,100 --> 00:02:48,660
was dice rolls.
40
00:02:49,440 --> 00:02:53,340
Now the reaction is weighing your hardware wallet
41
00:02:53,340 --> 00:02:57,220
on a scale, asking for clear cases.
42
00:02:57,480 --> 00:03:00,140
I mean, folks are obviously very concerned
43
00:03:00,140 --> 00:03:02,720
and so I'm interested from the perspective
44
00:03:02,720 --> 00:03:04,900
of obviously a company making these devices
45
00:03:04,900 --> 00:03:07,620
and trying to put in more advanced features
46
00:03:07,620 --> 00:03:09,000
like active anti-tamper
47
00:03:09,000 --> 00:03:10,960
that doesn't exist in too many devices.
48
00:03:11,560 --> 00:03:13,560
But I'm also interested in helping,
49
00:03:14,020 --> 00:03:18,300
you know, helping everyone in the industry
50
00:03:18,300 --> 00:03:20,040
kind of figure out the right path forward
51
00:03:20,040 --> 00:03:21,960
and not get scared of self-custody
52
00:03:21,960 --> 00:03:24,440
because there's a lot of things that we can do here.
53
00:03:24,700 --> 00:03:27,000
So obviously I have a knowledge of all of this.
54
00:03:27,100 --> 00:03:31,080
I have a knowledge of that specific Nordic chip as well,
55
00:03:31,080 --> 00:03:36,040
that LTE modem that Mark found in these devices.
56
00:03:36,820 --> 00:03:40,640
We were looking at that part for a future product at Foundation.
57
00:03:40,900 --> 00:03:45,840
So as soon as I saw that, I was asking Mark about things like, you know,
58
00:03:45,880 --> 00:03:48,260
the GPS, right, and that antenna.
59
00:03:48,580 --> 00:03:53,840
And we were speculating about as well that this thing could be more compact
60
00:03:53,840 --> 00:03:54,420
in the future.
61
00:03:54,520 --> 00:03:58,280
So I'm more than happy that we can kind of rob however you want to do it,
62
00:03:58,340 --> 00:04:00,240
dive in, I think, to all these different areas.
63
00:04:00,240 --> 00:04:16,500
But I was just saying to you before we started recording, it's a shame that I'm only coming on here within this solemn kind of mood with some another horrible event that's happened that people have lost like on the order of $100 million.
64
00:04:17,260 --> 00:04:18,220
It's just horrible.
65
00:04:19,440 --> 00:04:19,700
Yeah.
66
00:04:19,880 --> 00:04:21,480
Oh, please, Mark.
67
00:04:21,900 --> 00:04:22,080
Yeah.
68
00:04:22,160 --> 00:04:22,380
Sorry.
69
00:04:22,500 --> 00:04:27,520
I think I read
Load full transcript
recently that actually the estimated loss has gone over $100 million recently.
70
00:04:29,020 --> 00:04:29,460
My God.
71
00:04:29,980 --> 00:04:31,820
So yeah, I mean, let's set the stage, right?
72
00:04:31,980 --> 00:04:35,340
We just saw people lose $100 million worth of crypto.
73
00:04:35,640 --> 00:04:36,420
How did it happen?
74
00:04:36,600 --> 00:04:40,860
They had hardware wallets that had malicious implants put inside of them.
75
00:04:40,860 --> 00:04:44,540
And then when they used them and sent their crypto to these wallets,
76
00:04:44,740 --> 00:04:48,240
they could then be easily drained because these wallets were transmitting
77
00:04:48,240 --> 00:04:51,340
those seed phrases out to the hackers.
78
00:04:51,840 --> 00:04:55,340
Okay, so Mark, let's get into the technicals of how this actually worked.
79
00:04:55,620 --> 00:04:57,920
You got the device delivered, you opened it up.
80
00:04:57,920 --> 00:05:00,580
What are the pieces that were not supposed to be there,
81
00:05:00,580 --> 00:05:03,420
and how did it send out the seed phrases?
82
00:05:03,420 --> 00:05:04,960
Well, when you get the device,
83
00:05:04,960 --> 00:05:21,977
basically it comes in a box and the first thing you do you look I mean you typically open it you see in the box just the ledger Most normal people would just stop here connect it you go to your computer your computer tells you the device is genuine
84
00:05:22,617 --> 00:05:27,577
you feel good about it, and you don't care much about what's actually inside the device.
85
00:05:29,077 --> 00:05:34,197
But this specific attack, there's absolutely nothing you can see from outside.
86
00:05:34,197 --> 00:05:37,677
you need to actually open the device to find it.
87
00:05:38,377 --> 00:05:41,437
For the Nano X, it's actually fairly easy to open.
88
00:05:41,937 --> 00:05:45,297
The other models like the Stacks, I think, and the Flex
89
00:05:45,297 --> 00:05:46,697
are much more difficult to open.
90
00:05:46,817 --> 00:05:49,837
I'm actually going to be making a video
91
00:05:49,837 --> 00:05:52,477
about opening each of those after we're finishing.
92
00:05:53,357 --> 00:05:55,937
But what you're expected to find inside your ledger
93
00:05:55,937 --> 00:05:59,597
is the board, which is very small.
94
00:05:59,597 --> 00:06:03,477
And normally, you only have a few things on it.
95
00:06:04,197 --> 00:06:13,617
On one side, you've got, let me show this, the secure element, which is right here, and you've got the main chip that handles USB and Bluetooth.
96
00:06:14,437 --> 00:06:23,677
And on the other side, you've got this side where you have the connector for the screen and the number of little copper circles, which are test points.
97
00:06:24,117 --> 00:06:27,937
And normally those test points, there shouldn't be anything on any of these.
98
00:06:27,937 --> 00:06:31,697
and it turns out that the few ones that are right on this side,
99
00:06:32,017 --> 00:06:33,457
which is where the screen connects,
100
00:06:34,097 --> 00:06:36,337
do contain the information that is shown on the screen
101
00:06:36,337 --> 00:06:44,417
and this is what the SPY implant will try to connect to and get data from.
102
00:06:44,777 --> 00:06:48,497
So the SPY implant, it would bypass the secure element
103
00:06:48,497 --> 00:06:51,817
and just read directly what was being put on the screen.
104
00:06:51,997 --> 00:06:54,017
Zach, if you want to describe it a little bit more, please.
105
00:06:54,497 --> 00:06:56,037
Yeah, I mean, it's kind of genius, right?
106
00:06:56,037 --> 00:07:10,557
And as soon as Mark posted about this, before I saw Mark's subsequent post where he had explained how it's just reading the screen, I speculated that the only way to do this would be to read the screen.
107
00:07:10,557 --> 00:07:20,937
Because there was a bunch of, I would say, confusion initially where there was some tweet that went viral that claimed that the secure element was compromised, right?
108
00:07:20,957 --> 00:07:24,917
That they were able to read or exfiltrate data from the secure element.
109
00:07:24,917 --> 00:07:39,477
And then I was thinking, how would I do this? And one thing that Ledger does, right, is it shows the seed words it's set up. Now, the newer devices do not necessarily do that because they offer an NFC card type backup option.
110
00:07:39,477 --> 00:07:46,317
kind of we offer a different kind of option like that but it's genius because you can have this
111
00:07:46,317 --> 00:07:53,397
incredibly secure device with this secure element chip that that you know ledger will will claim
112
00:07:53,397 --> 00:07:59,137
that it's impenetrable right and then one of the big marketing things that ledger does is they talk
113
00:07:59,137 --> 00:08:04,697
about how they have the secure element drives the display directly and and that's one of their big
114
00:08:04,697 --> 00:08:09,977
marketing claims on the e-ink models, right? The stacks and the flux, which is actually a really
115
00:08:09,977 --> 00:08:15,097
great idea. I think they chose e-ink so that it can be driven directly by the secure element,
116
00:08:15,097 --> 00:08:20,337
which would not be able to drive a more high resolution, you know, color type of display.
117
00:08:20,797 --> 00:08:26,557
But in my quick research, and I was actually, even as a, you know, a maker of these types of devices,
118
00:08:26,557 --> 00:08:33,357
I was surprised to see that there does not even exist the ability to encrypt signals
119
00:08:33,357 --> 00:08:35,377
that go to a display.
120
00:08:35,597 --> 00:08:38,297
Even in the most hardcore situations,
121
00:08:38,697 --> 00:08:41,897
there's no, like if we want to go procure a display,
122
00:08:42,617 --> 00:08:45,657
as soon as like, there's a data connector, right?
123
00:08:45,757 --> 00:08:47,397
There's, you know, a ribbon cable,
124
00:08:47,517 --> 00:08:49,677
like this is the one that we use for Passport Prime.
125
00:08:50,177
--> 00:08:53,677
You plug into this, this is just sending data
126
00:08:53,677 --> 00:08:55,877
that's going to be converted into raw pixels.
127
00:08:56,397 --> 00:08:57,317
And that's it.
128
00:08:57,417 --> 00:09:00,097
And so if you can tap into this, you know, cable,
129
00:09:00,097 --> 00:09:02,337
or in this case, the different cables
130
00:09:02,337 --> 00:09:03,817
that come on these ledger devices,
131
00:09:03,817 --> 00:09:08,017
you can effectively read the data being displayed
132
00:09:08,017 --> 00:09:09,197
on the screen.
133
00:09:09,317 --> 00:09:11,977
And now there's a question of reading that,
134
00:09:12,097 --> 00:09:12,557
converting that.
135
00:09:12,697 --> 00:09:14,297
I believe Mark could probably give us more insight
136
00:09:14,297 --> 00:09:14,717
into there.
137
00:09:14,817 --> 00:09:16,137
You said something about like the fonts
138
00:09:16,137 --> 00:09:17,657
they were specifically looking at,
139
00:09:17,977 --> 00:09:18,957
like the fonts used.
140
00:09:19,077 --> 00:09:20,457
And it was a little more complex.
141
00:09:20,697 --> 00:09:23,837
And I think what was most shocking to me
142
00:09:23,837 --> 00:09:26,497
was how refined it felt.
143
00:09:26,617 --> 00:09:27,777
I'm curious what you think,
144
00:09:27,777 --> 00:09:31,157
but it feels like this was not a first generation.
145
00:09:31,157 --> 00:09:32,757
no this is definitely
146
00:09:32,757 --> 00:09:35,337
yeah I mean we found a number
147
00:09:35,337 --> 00:09:37,317
of implants there's a
148
00:09:37,317 --> 00:09:39,077
slightly older one which is
149
00:09:39,077 --> 00:09:41,037
on this fairly
150
00:09:41,037 --> 00:09:42,897
visible circuit
151
00:09:42,897 --> 00:09:44,777
the new ones actually
152
00:09:44,777 --> 00:09:47,077
I'm not sure I don't have a black
153
00:09:47,077 --> 00:09:49,117
but generally the ledger
154
00:09:49,117 --> 00:09:51,237
board used to be black and there's a lot of black
155
00:09:51,237 --> 00:09:53,357
boards out there and
156
00:09:53,357 --> 00:09:55,317
the new implant is made to
157
00:09:55,317 --> 00:09:57,317
be black and this
158
00:09:57,317 --> 00:09:58,277
is extremely
159
00:09:58,277 --> 00:10:01,017
well made
160
00:10:01,017 --> 00:10:03,177
because if you look here, I don't know if you can see,
161
00:10:03,557 --> 00:10:06,617
there are a few holes right around this part.
162
00:10:07,517 --> 00:10:10,857
Those holes are exactly where the video signal goes.
163
00:10:11,297 --> 00:10:12,517
So you've got your ledger board.
164
00:10:13,237 --> 00:10:17,797
You can just clip it, and it just clips into place,
165
00:10:18,257 --> 00:10:31,093
and then you just have three solder points to do and the spy implant can see everything that happens on the screen And on the software side it will see the screen it will detect when you are ready and your seed is being
166
00:10:31,093 --> 00:10:37,373
shown on the screen and it will actually OCR so optically recognize the words of your seed
167
00:10:37,373 --> 00:10:43,433
as they are shown on the screen put this into a buffer and once all your seed has been shown
168
00:10:43,433 --> 00:10:46,113
it will just basically
169
00:10:46,113 --> 00:10:48,173
phone home. It's actually
170
00:10:48,173 --> 00:10:49,653
a code because it's using the data
171
00:10:49,653 --> 00:10:52,173
and it will send the seed over
172
00:10:52,173 --> 00:10:53,553
to the attacker server.
173
00:10:54,313 --> 00:10:55,733
And I just want to say Rob too
174
00:10:55,733 --> 00:10:58,113
the most striking thing about this is I've
175
00:10:58,113 --> 00:11:00,153
always been a big fan of passphrase
176
00:11:00,153 --> 00:11:02,273
you know in addition to seed
177
00:11:02,273 --> 00:11:04,453
and I mean I'm agnostic
178
00:11:04,453 --> 00:11:06,553
I would say to multi-seg historically
179
00:11:06,553 --> 00:11:08,393
I say you know it's good for some cases
180
00:11:08,393 --> 00:11:10,213
it can be complicated for other cases
181
00:11:10,213 --> 00:11:12,433
but I don't know Mark if you've
182
00:11:12,433 --> 00:11:18,913
confirmed but theoretically i mean anytime you enter a passphrase it's also displayed on device
183
00:11:18,913 --> 00:11:25,353
and and you have to do that yeah yeah anything that's shown on the screen uh the implant can see
184
00:11:25,353 --> 00:11:31,173
and so for the people who are relatively new to bitcoin and crypto why does it matter that it can
185
00:11:31,173 --> 00:11:36,553
read the screen it's because when you first load up a hardware wallet it gives you your 12 or 24
186
00:11:36,553 --> 00:11:42,953
words. That is your seed phrase. That allows you to move your Bitcoin at any time to any other
187
00:11:42,953 --> 00:11:47,973
wallet. So if you have those 12 words, if the attacker has those 12 words, they might as well
188
00:11:47,973 --> 00:11:54,693
have your Bitcoin, right? So if they could read it off the device, then it's theirs. And the most
189
00:11:54,693 --> 00:11:58,933
important thing is that they could send it. So talk to us about sending the data. Like they
190
00:11:58,933 --> 00:12:02,813
literally had a chip in there that was as good as a cell phone. Like how did that work?
191
00:12:03,393 --> 00:12:06,113
Well, there's actually one thing that's very important
192
00:12:06,113 --> 00:12:10,033
is your seed phrase is basically everything that protects your Bitcoins.
193
00:12:10,913 --> 00:12:12,973
And the fact that it's been given to the attacker,
194
00:12:13,233 --> 00:12:14,333
the seed is basically worthless,
195
00:12:14,653 --> 00:12:18,433
which is kind of the same as what happened with cold card.
196
00:12:18,433 --> 00:12:19,993
If the attacker can know the seed,
197
00:12:20,513 --> 00:12:25,993
be it through a weak RNG or by getting access to your seed,
198
00:12:26,753 --> 00:12:30,893
what happens then is the hacker can see how much Bitcoin you have.
199
00:12:30,893 --> 00:12:34,253
they can monitor any crypto you
receive afterward,
200
00:12:34,833 --> 00:12:38,513
and they can actually send any of your crypto at any time.
201
00:12:39,553 --> 00:12:40,673
And in this case,
202
00:12:41,193 --> 00:12:45,893
part of the components that are included inside the...
203
00:12:47,653 --> 00:12:52,893
With the SPICE implant,
204
00:12:53,833 --> 00:12:57,093
you have specifically this larger tip,
205
00:12:57,093 --> 00:13:02,893
which is a LTE IoT chip.
206
00:13:03,493 --> 00:13:04,973
So there are chips that are made for IoT
207
00:13:04,973 --> 00:13:06,593
that are very low consumption
208
00:13:06,593 --> 00:13:09,413
and can be configured to only activate
209
00:13:09,413 --> 00:13:12,253
when, for example, you develop an IoT device
210
00:13:12,253 --> 00:13:13,973
it needs to communicate.
211
00:13:14,553 --> 00:13:18,033
And it uses a protocol that is itself made for IoT
212
00:13:18,033 --> 00:13:20,413
that allows it to just connect quickly,
213
00:13:20,613 --> 00:13:22,013
send something, receive something,
214
00:13:22,333 --> 00:13:23,233
and close the connection.
215
00:13:23,933 --> 00:13:26,633
And this is nice if you have an IoT device,
216
00:13:26,633 --> 00:13:33,013
for example, an alarm or something that needs to phone them to send you a notification on
217
00:13:33,013 --> 00:13:35,013
your phone about something that just happened.
218
00:13:35,693 --> 00:13:41,813
But for this ledger, what it's doing is it's actually calling the hacker and telling, hey,
219
00:13:41,933 --> 00:13:42,793
I've got a new seed.
220
00:13:42,933 --> 00:13:43,613
Here is a seed.
221
00:13:44,493 --> 00:13:50,113
So this happened just once as you set up your seed and maybe during the password setup later.
222
00:13:50,113 --> 00:13:53,233
but as soon as a hacker has the seed
223
00:13:53,233 --> 00:13:56,973
that can at any time in the future
224
00:13:56,973 --> 00:13:59,253
wait, see that you receive some crypto
225
00:13:59,253 --> 00:14:01,973
maybe wait a little more to see if you're going to add more
226
00:14:01,973 --> 00:14:04,653
and at any time the hacker can just say
227
00:14:04,653 --> 00:14:06,593
okay I'm going to just take everything
228
00:14:06,593 --> 00:14:09,853
yeah if I could add too
229
00:14:09,853 --> 00:14:11,953
I mean this IoT
230
00:14:11,953 --> 00:14:14,153
Internet of Things style chip
231
00:14:14,153 --> 00:14:16,153
it uses LTE network
232
00:14:16,153 --> 00:14:19,313
so 4G you might call that not 5G
233
00:14:19,313 --> 00:14:25,913
it's designed for smaller amounts of data so it cannot transmit gigabytes of data we don't need
234
00:14:25,913 --> 00:14:32,333
to transmit gigabytes it it handles kilobytes of data perfectly and so when you're just sending
235
00:14:32,333 --> 00:14:37,813
texts right like seed words and a little bit of information about that it handles it perfectly
236
00:14:37,813 --> 00:14:47,813
additionally um there are many companies that will sell global iot e-sim or just normal sim
237
00:14:47,813
--> 00:14:57,073
but in this case, eSIM data plans, where you could even buy data plans for a pool of eSIMs.
238
00:14:57,153 --> 00:15:05,413
So depending on how many of these chips are out there, this data was purchased somewhere with some IoT provider.
239
00:15:05,753 --> 00:15:11,973
It may have been done individually, but it can also be done as a pool where you just pay a certain monthly fee,
240
00:15:12,193 --> 00:15:17,113
typically per eSIM, and then you pool the data for these IoT plans.
241
00:15:17,113 --> 00:15:23,273
what's really great about this chip is that it's global i believe they have a bunch of different
242
00:15:23,273 --> 00:15:28,093
variants but typically with lte chips like you may have issues where it doesn't work in every
243
00:15:28,093 --> 00:15:44,150
country right you you buy a chip for one region you know it only works in north america you need a different one for you know eu or something like that but in this case with this Nordic Semi which is just it a company it a very well regarded company in
244
00:15:44,150 --> 00:15:48,670
the space that makes these types of chips. There's others as well. I'm curious if we'll see other
245
00:15:48,670 --> 00:15:55,670
implants made by other companies, but this one is a very refined chip. And you can get this kind of
246
00:15:55,670 --> 00:16:01,610
global data plan for it. So no matter where you are in the world, right, it'll just send that in.
247
00:16:01,610 --> 00:16:05,230
And also, I believe these plans, they can be purchased as a company.
248
00:16:05,730 --> 00:16:06,930
I'm very curious, right?
249
00:16:07,130 --> 00:16:09,030
Are there shell companies?
250
00:16:09,350 --> 00:16:13,330
Like, will you ever, even if you know who, if the provider knows that they sold this plan,
251
00:16:13,430 --> 00:16:16,910
like, would we ever even figure out who actually owned these?
252
00:16:16,990 --> 00:16:21,310
Because these are for companies that might be deploying thousands of IoT devices.
253
00:16:21,510 --> 00:16:25,750
Imagine you're like farming and you need to have sensors all over the place.
254
00:16:25,890 --> 00:16:31,510
That's the kind of use case for these kinds of devices and these kinds of IoT cellular plans.
255
00:16:31,610 --> 00:16:34,650
another place where you have a lot of those
256
00:16:34,650 --> 00:16:36,210
IoT chips is inside cars
257
00:16:36,210 --> 00:16:38,950
typical cars
258
00:16:38,950 --> 00:16:40,470
nowadays has at least
259
00:16:40,470 --> 00:16:42,430
two of those data plans
260
00:16:42,430 --> 00:16:44,530
one is typically used by the
261
00:16:44,530 --> 00:16:46,650
navigation system and the second one is used
262
00:16:46,650 --> 00:16:48,070
by the telemetry system
263
00:16:48,070 --> 00:16:50,730
that will send information
264
00:16:50,730 --> 00:16:52,210
about your car to the maker
265
00:16:52,210 --> 00:16:54,210
like a GPS location and so on
266
00:16:54,210 --> 00:16:55,270
so
267
00:16:55,270 --> 00:16:58,290
you also have those in flock cameras
268
00:16:58,290 --> 00:16:59,050
in the US
269
00:16:59,050 --> 00:17:05,850
it's basically everywhere and the lte network is great because you can send a lot of data you can
270
00:17:05,850 --> 00:17:11,770
send a little bit of data a lot of makers have been using this uh for those implants i believe
271
00:17:11,770 --> 00:17:21,130
they are actually uh taking lte chips from old uh e-waste and just recycling them uh for this use
272
00:17:21,130 --> 00:17:27,130
all right so that's the technical aspect now let's get into the timeline of how this all played out so
273
00:17:27,130 --> 00:17:35,750
So it was last Thursday, October 8th, you put out a postmark saying that, hey, if you're looking for an eSIM card, just look inside the ledger.
274
00:17:35,950 --> 00:17:37,550
Right. So that was sort of tongue in cheek.
275
00:17:37,750 --> 00:17:41,950
Did you know that you had stumbled upon some massive attack when you posted that?
276
00:17:42,090 --> 00:17:48,270
Well, actually, three weeks before I already posted about this when I found the implant.
277
00:17:48,470 --> 00:17:50,110
But at the time, nobody really paid attention.
278
00:17:50,110 --> 00:17:57,450
and for me being on the more technical side the implant was very advanced in terms of technology
279
00:17:57,450 --> 00:18:03,350
so I did believe it was consequent so I actually tagged Ledger I tried to contact some people
280
00:18:03,350 --> 00:18:07,890
I didn't get any response at the time and three weeks later someone talks about
281
00:18:07,890 --> 00:18:16,230
taking eSIMs out of lock cameras as a free data plan and that would be very illegal but at the
282
00:18:16,230 --> 00:18:22,930
same time I have an already illegal device that has an eSIM. So it's, I said it was a nice joke
283
00:18:22,930 --> 00:18:28,810
to just say, if you want to take an eSIM, at least take it from something like the ledger,
284
00:18:28,890 --> 00:18:33,910
you can buy a very cheap one. If you remove the implant, it's actually cheaper than a real ledger
285
00:18:33,910 --> 00:18:40,210
and you get a free eSIM as a bonus. So you're making a joke after not really being respected
286
00:18:40,210 --> 00:18:42,750
for calling out this problem the weeks before.
287
00:18:43,010 --> 00:18:46,570
And then the next day, reports on X and Reddit
288
00:18:46,570 --> 00:18:51,850
of over $86 million of crypto being drained out of these wallets,
289
00:18:51,970 --> 00:18:54,230
$70 million of Tether on Tron,
290
00:18:54,750 --> 00:18:58,750
I think around $15, $60 million worth of Bitcoin.
291
00:18:59,350 --> 00:19:01,230
So these are serious numbers.
292
00:19:01,970 --> 00:19:04,290
Do you think that your post getting traction
293
00:19:04,290 --> 00:19:07,290
is what set off the hackers to drain the wallets?
294
00:19:07,910 --> 00:19:08,690
I don't know.
295
00:19:08,690 --> 00:19:16,050
mean, I guess that could be possible. And if that's the case,
I'm happy I was able to,
296
00:19:16,770 --> 00:19:22,230
in a way, make this stuff before it got worse. But really, there's no way to know until we,
297
00:19:22,490 --> 00:19:27,690
I mean, those guys get caught. Yeah. And so then we're going to dive into the operation that was
298
00:19:27,690 --> 00:19:32,090
going on. But Zach, on the timeline, when you first heard this, what were you thinking?
299
00:19:32,330 --> 00:19:38,230
I didn't think it would be as widespread as it actually was. I just, you know, I assumed that
300
00:19:38,230 --> 00:19:41,570
there was a small number or something like that.
301
00:19:41,690 --> 00:19:45,570
I was blown away by the scale of this operation.
302
00:19:45,710 --> 00:19:47,350
I think that's what I keep coming back to,
303
00:19:47,970 --> 00:19:50,290
the scale and professionalism of this,
304
00:19:50,410 --> 00:19:52,930
the fact that there were multiple generations of it.
305
00:19:54,230 --> 00:19:56,490
And then also I was thinking more about
306
00:19:56,490 --> 00:19:59,070
comparing this to the cold card incident.
307
00:19:59,070 --> 00:20:04,050
Like how do you responsibly disclose this?
308
00:20:05,010 --> 00:20:07,630
You know, what's the right option for this to happen?
309
00:20:08,230 --> 00:20:13,190
I think Mark probably did the right thing by just sharing this in real time.
310
00:20:15,090 --> 00:20:22,650
I believe some reports are coming out now that Ledger may have been aware of this earlier in the year, I saw.
311
00:20:22,850 --> 00:20:25,050
I don't know, Mark, if you had seen some of those posts.
312
00:20:25,910 --> 00:20:26,070
Yes.
313
00:20:27,290 --> 00:20:32,050
Actually, there was someone who posted about this and asked about it.
314
00:20:32,630 --> 00:20:36,130
He said he had a video call with Ledger end of last year.
315
00:20:36,130 --> 00:20:49,990
So this is, yeah, so just the point I'm trying to make is this is a really difficult situation and it makes me think back to the cold card incident where we were thinking there were some thought experiments, right? You know, what would happen?
316
00:20:50,011 --> 00:21:01,971
If you discovered this and you were the only one in the world who knew that these cold card Mark III devices had, you know, this entropy bug, would you just announce it?
317
00:21:01,971 --> 00:21:07,151
Would you try to get as many people as possible to move their funds as fast as possible?
318
00:21:07,151 --> 00:21:14,211
would you responsibly disclose and then maybe cold card could have like worked with bitcoin
319
00:21:14,211 --> 00:21:21,071
podcasters to make up some timeline of events to say there's some reason you need to migrate your
320
00:21:21,071 --> 00:21:27,631
funds but you know the second they would announce that the attacker would go and start sweeping the
321
00:21:27,631 --> 00:21:34,191
wallets and we it's so interesting to think about those what ifs with the cold card situation
322
00:21:34,191 --> 00:21:36,911
knowing that we only found out
323
00:21:36,911 --> 00:21:40,191
because attackers started sweeping wallets.
324
00:21:40,991 --> 00:21:42,131
And in this situation,
325
00:21:42,131 --> 00:21:44,331
we now have some confirmation
326
00:21:44,331 --> 00:21:48,731
that Ledger was notified some number of months ago
327
00:21:48,731 --> 00:21:51,071
and no announcement or disclosure
328
00:21:51,071 --> 00:21:52,951
was made to the customer base.
329
00:21:53,051 --> 00:21:55,231
And so it's just really interesting.
330
00:21:55,231 --> 00:21:57,031
And when I say interesting,
331
00:21:57,151 --> 00:21:58,051
I also mean kind of horrible
332
00:21:58,051 --> 00:21:59,911
because I think, well, what if it happened to us
333
00:21:59,911 --> 00:22:00,991
at Foundation?
334
00:22:01,551 --> 00:22:02,591
What would we do?
335
00:22:02,591 --> 00:22:06,791
and I think probably Ledger is grappling with this idea of
336
00:22:06,791 --> 00:22:09,471
you don't want to cause mass panic either.
337
00:22:09,971 --> 00:22:12,831
You don't want to cause a majority of these customers
338
00:22:12,831 --> 00:22:15,211
to end up losing money in a panic
339
00:22:15,211 --> 00:22:17,851
because they are moving their funds,
340
00:22:18,011 --> 00:22:19,191
they don't know what they're doing,
341
00:22:19,811 --> 00:22:22,951
they make a mistake and now they don't have access to their money.
342
00:22:23,151 --> 00:22:25,691
So I think that's how I'm thinking about it,
343
00:22:25,691 --> 00:22:27,471
not necessarily just about this event,
344
00:22:27,631 --> 00:22:29,611
but these kinds of what-ifs
345
00:22:29,611 --> 00:22:33,671
And, you know, what is the responsibility of the person who discovers this?
346
00:22:33,811 --> 00:22:35,491
What is the responsibility of the company?
347
00:22:36,151 --> 00:22:37,851
And it's kind of a mess right now.
348
00:22:37,891 --> 00:22:43,071
And we've seen almost both of these scenarios now play out over a span of like two months.
349
00:22:44,311 --> 00:22:47,971
If you're anything like me, the cold card hack shook you to the core.
350
00:22:48,591 --> 00:22:51,931
Personally, it made me completely rethink how I custody my Bitcoin.
351
00:22:52,711 --> 00:22:54,611
My biggest takeaway was pretty simple.
352
00:22:54,911 --> 00:22:58,271
I don't want all of my Bitcoin relying on a single custody model.
353
00:22:58,271 --> 00:23:03,271
For me, diversifying how I secure my Bitcoin is what helps me sleep better at night.
354
00:23:04,031 --> 00:23:05,471
That's what led me to OnRamp.
355
00:23:05,651 --> 00:23:10,971
OnRamp offers multi-institution custody, where three independent institutions protect your Bitcoin,
356
00:23:11,331 --> 00:23:13,611
none of which can move it, use it, or lose it.
357
00:23:14,471 --> 00:23:17,771
OnRamp also maintains custody insurance through Lloyds of London.
358
00:23:18,551 --> 00:23:21,731
And beyond custody, they offer a full suite of financial services,
359
00:23:22,071 --> 00:23:25,131
including the ability to buy Bitcoin directly to cold storage.
360
00:23:25,131 --> 00:23:29,111
But the thing that impressed me the most about OnRamp is how they handle inheritance.
361
00:23:29,431 --> 00:23:33,191
If something happened to me, the last thing I'd want is for my wife to have to go on a treasure hunt,
362
00:23:33,331 --> 00:23:35,151
trying to figure out how to access our Bitcoin.
363
00:23:36,171 --> 00:23:42,011
With OnRamp, you can designate your spouse or other beneficiaries as part of your inheritance plan from the very beginning.
364
00:23:42,271 --> 00:23:46,991
It takes something that can be incredibly complicated with Bitcoin and makes it much easier to plan for.
365
00:23:47,111 --> 00:23:51,691
That's why OnRamp has become part of how I think about protecting my family's Bitcoin for generations.
366
00:23:51,691 --> 00:23:57,691
Go to onrampbitcoin.com to learn more and book your complimentary consultation today.
367
00:24:00,491 --> 00:24:05,231
And if we maintain with the timeline, it then came out on Saturday.
368
00:24:05,611 --> 00:24:09,471
What was the source of all these ledgers that were maliciously attacked?
369
00:24:09,771 --> 00:24:15,151
Well, it turns out that there was a official ledger reseller in Malaysia called Crypto Billis.
370
00:24:15,411 --> 00:24:17,011
These guys had been around for a couple of years.
371
00:24:17,111 --> 00:24:17,831
They were respected.
372
00:24:18,271 --> 00:24:19,831
They had sold ledgers for years, right?
373
00:24:19,831 --> 00:24:26,071
Well, back in March of this year, they sold the company to someone else in China, we believe.
374
00:24:26,631 --> 00:24:33,831
And they signed an NDA with the new owner that said you cannot disclose the fact that you sold your business until October 19th.
375
00:24:34,791 --> 00:24:36,711
Oh, pretty crazy how all this all lines up.
376
00:24:36,771 --> 00:24:37,411
The timeline is.
377
00:24:37,771 --> 00:24:42,291
So they sold the company to these other guys in China, I guess.
378
00:24:42,371 --> 00:24:45,651
And then it seems as if that's when this attack occurred.
379
00:24:45,651 --> 00:24:51,431
They used the good graces, the good name of Cryptobilist to sell these products to the Asian market.
380
00:24:51,991 --> 00:24:56,011
And Ledger confirms that, yes, this is the source of these chips that Mark has found.
381
00:24:56,711 --> 00:24:59,291
So talk to us about that.
382
00:24:59,851 --> 00:25:02,871
Resellers, malicious actors buying these companies.
383
00:25:03,371 --> 00:25:05,351
Mark, have you ever heard of anything like this before?
384
00:25:06,271 --> 00:25:08,971
No, that was the first time I've heard about this kind of operation.
385
00:25:08,971 --> 00:25:21,011
And just like Zach said, it's really a question of scale because you can imagine maybe someone buying a ledger, adding an implant, returning it, hoping that someone else would receive it.
386
00:25:21,331 --> 00:25:27,111
But buying the actual shop where the ledger is being sold is unheard of.
387
00:25:27,191 --> 00:25:30,611
It's the first time we ever got these kind of things happening.
388
00:25:30,611 --> 00:25:32,951
and I don't know
389
00:25:32,951 --> 00:25:34,991
for how much crypto bill is sold
390
00:25:34,991 --> 00:25:37,131
but
391
00:25:37,131 --> 00:25:39,171
the scale is really
392
00:25:39,171 --> 00:25:41,291
impressive
393
00:25:41,291 --> 00:25:43,051
here and I would guess that
394
00:25:43,051 --> 00:25:44,051
based on the
395
00:25:44,051 --> 00:25:47,291
returns they've made basically that was worth it
396
00:25:47,291 --> 00:25:47,531
but
397
00:25:47,531 --> 00:25:50,051
that takes
398
00:25:50,051 --> 00:25:50,991
a
399
00:25:50,991 --> 00:25:54,751
very large operation to do that right
400
00:25:54,751 --> 00:26:11,145
buying a shop probably as I probably sent some people to audit the inventory to actually install the implants Probably a lot of the thing we don know yet in the operation that was involved here
401
00:26:12,625 --> 00:26:19,965
But overall, this is something we see for the first time, so we don't know yet how to deal with it.
402
00:26:20,625 --> 00:26:25,525
For me, the most important was to make sure that as many people as possible would know about this
403
00:26:25,525 --> 00:26:26,945
and know how to check their own devices,
404
00:26:27,125 --> 00:26:28,985
because that's the only way we would catch this.
405
00:26:29,905 --> 00:26:33,045
And also I wanted to make sure that there was nothing similar
406
00:26:33,045 --> 00:26:34,445
happening in other parts of the world,
407
00:26:34,945 --> 00:26:37,305
because if it's done once,
408
00:26:37,645 --> 00:26:40,085
there could be also more of this in other places.
409
00:26:40,085 --> 00:26:43,745
And as I said, this implant is built to be easy to install,
410
00:26:43,885 --> 00:26:47,525
which points to basically a mass production
411
00:26:47,525 --> 00:26:49,205
and mass distribution of it,
412
00:26:49,905 --> 00:26:53,785
which for me would mean that it's likely not limited
413
00:26:53,785 --> 00:26:55,305
to just one location.
414
00:26:55,525 --> 00:27:04,565
Yeah. And Zach, I mean, you run a hardware wallet business. Do you deal with resellers? Do you vet them? What is that relationship like?
415
00:27:04,565 --> 00:27:16,425
we do and we don't deal with too many we we we call it now value added resellers right where we
416
00:27:16,425 --> 00:27:23,045
don't just want to blast them out everywhere we want to make sure we know who they are but i mean
417
00:27:23,045 --> 00:27:27,765
we don't really know who that how could we possibly know who they are i mean look at what
418
00:27:27,765 --> 00:27:33,425
just happened so we're currently evaluating what steps we want to take on the reseller side of
419
00:27:33,425 --> 00:27:40,465
things. And I think we also used to sell on Amazon for a period of time and we stopped. It made me
420
00:27:40,465 --> 00:27:46,665
very uncomfortable. I also
know that Amazon is one of the routes where you can get counterfeit
421
00:27:46,665 --> 00:27:53,025
devices or I should say product in general. The supplement industry has been struggling with that
422
00:27:53,025 --> 00:28:00,085
for years where you can sell product on Amazon and it will show up under like the umbrella listing
423
00:28:00,085 --> 00:28:02,185
for that item, even through different sellers.
424
00:28:02,505 --> 00:28:04,125
And so we stopped that.
425
00:28:05,045 --> 00:28:08,225
But our devices now have,
426
00:28:08,425 --> 00:28:11,845
our Passport Prime device has an active anti-tamper feature
427
00:28:11,845 --> 00:28:14,665
where if you try to break it open, it will wipe itself.
428
00:28:15,205 --> 00:28:17,565
But we haven't done a good enough job
429
00:28:17,565 --> 00:28:19,805
on the supply chain attack side of things.
430
00:28:19,965 --> 00:28:21,465
We're beefing that up right now,
431
00:28:22,125 --> 00:28:26,245
where if anyone would open it up
432
00:28:26,245 --> 00:28:27,705
before it arrives to the customer,
433
00:28:27,705 --> 00:28:31,685
it would fail the initial supply chain validation.
434
00:28:33,065 --> 00:28:37,025
But even then it still makes us very concerned about resellers
435
00:28:37,025 --> 00:28:40,605
because you could start to play out all these different permutations
436
00:28:40,605 --> 00:28:42,665
of these attacks.
437
00:28:42,665 --> 00:28:47,985
The easiest one would be to swap all the internals of the device
438
00:28:47,985 --> 00:28:50,705
with something malicious and trick our customers
439
00:28:50,705 --> 00:28:54,085
into downloading an app or using a web app that was malicious
440
00:28:54,085 --> 00:28:57,365
and then it would completely cut us out of that process
441
00:28:57,365 --> 00:29:10,365
And so I think, like I said, we're still trying to figure out, but I think the place we might end up is just you buy from us, you ship it directly from our American factory to the customer.
442
00:29:11,185 --> 00:29:12,825
And that's that, right?
443
00:29:12,885 --> 00:29:17,245
And we have great supply chain validation and anti-tamper features baked into the hardware.
444
00:29:17,525 --> 00:29:21,025
And it's just really difficult right now.
445
00:29:21,685 --> 00:29:24,745
I wish I had, you know, more clear, better answers.
446
00:29:24,745 --> 00:29:34,025
but I do think that I think that this idea that you would you would have a distributors and
447
00:29:34,025 --> 00:29:38,445
resellers and everything like that and you you lose that relationship with the customer
448
00:29:38,445 --> 00:29:44,325
I think really starts to break down in this industry you know I think I think you some of
449
00:29:44,325 --> 00:29:50,785
my crazy ideas this weekend were you know we if customers only bought from us we could pre-provision
450
00:29:50,785 --> 00:29:54,905
like a code, like an eight digit code onto the device.
451
00:29:55,325 --> 00:29:59,325
And like when you get it, the device is locked to you, right?
452
00:29:59,365 --> 00:30:03,645
And you type in that code and we could do, you can only do those kinds of things if you
453
00:30:03,645 --> 00:30:04,645
cut out the resellers.
454
00:30:04,785 --> 00:30:08,825
At the same time, we have some amazing, you know, resellers that we work with across the
455
00:30:08,825 --> 00:30:09,045
world.
456
00:30:09,165 --> 00:30:11,045
So we're grappling with that right now.
457
00:30:11,045 --> 00:30:16,985
But I worry that the whole reseller market is going to be kind of dead for hardware while
458
00:30:16,985 --> 00:30:17,785
it's moving forward.
459
00:30:18,405 --> 00:30:19,085
It changes everything.
460
00:30:19,225 --> 00:30:19,805
It really does.
461
00:30:19,805 --> 00:30:22,685
but this is a question for Mark.
462
00:30:22,745 --> 00:30:24,405
I mean, you've been around Bitcoin and crypto
463
00:30:24,405 --> 00:30:25,705
for a decade and a half.
464
00:30:25,945 --> 00:30:28,665
And so these guys stole the funds, right?
465
00:30:28,885 --> 00:30:30,565
And now they're trying to move it on chain.
466
00:30:30,685 --> 00:30:33,365
And so they sent the ETH to Tornado Cash
467
00:30:33,365 --> 00:30:34,945
and then into Zcash.
468
00:30:35,005 --> 00:30:37,545
And now it's out in the form of USDC.
469
00:30:37,825 --> 00:30:40,365
But there's some blockchain researchers
470
00:30:40,365 --> 00:30:42,205
who say they can track that exactly
471
00:30:42,205 --> 00:30:44,785
through all those ways to try to
472
00:30:44,785 --> 00:30:47,005
obfuscate where the funds are going, right?
473
00:30:47,225 --> 00:30:48,885
I still don't think that Bitcoin has moved.
474
00:30:48,885 --> 00:30:54,965
and I think Tether has frozen $10 million of the Tether that was part of this hack.
475
00:30:55,405 --> 00:30:59,325
So do you think these guys have any hope of using these funds,
476
00:30:59,585 --> 00:31:01,185
moving them or hiding them effectively?
477
00:31:01,785 --> 00:31:08,525
I mean, one of the issues with cryptocurrencies is that it's built to announce privacy,
478
00:31:08,745 --> 00:31:24,778
which is in a way a good thing but at the same time it can be used by bad actors to move and hide funds And it will be difficult for them to properly hide moving million or whatever they
479
00:31:24,778 --> 00:31:25,358
trying to move.
480
00:31:26,118 --> 00:31:31,298
Because even if, for example, you move this all in ton of cash, one day you have $50 million
481
00:31:31,298 --> 00:31:37,358
in, the other day you have $50 million out, it's still going to stand out quite a bit.
482
00:31:37,358 --> 00:31:41,258
I mean I'm not going to give any kind of advice
483
00:31:41,258 --> 00:31:42,658
to help this
484
00:31:42,658 --> 00:31:45,218
there are ways to track things
485
00:31:45,218 --> 00:31:46,098
and
486
00:31:46,098 --> 00:31:49,338
but we've
487
00:31:49,338 --> 00:31:51,538
seen actors, North Korea
488
00:31:51,538 --> 00:31:53,298
different managing to move
489
00:31:53,298 --> 00:31:54,378
much larger amounts
490
00:31:54,378 --> 00:31:57,138
in a way that
491
00:31:57,138 --> 00:31:59,338
at the end couldn't be tracked so I
492
00:31:59,338 --> 00:32:01,018
do believe that
493
00:32:01,018 --> 00:32:03,478
it will be very difficult to catch
494
00:32:03,478 --> 00:32:04,858
everything, the good thing is
495
00:32:04,858 --> 00:32:06,798
as you said USDT
496
00:32:06,798 --> 00:32:09,098
Teixeira has locked $10 million.
497
00:32:09,578 --> 00:32:13,878
I think there was another smaller amount,
498
00:32:14,038 --> 00:32:15,018
but still a few million dollars
499
00:32:15,018 --> 00:32:15,998
that were blocked somewhere else.
500
00:32:16,738 --> 00:32:17,898
But even for those amounts,
501
00:32:17,978 --> 00:32:19,178
it's going to be very difficult
502
00:32:19,178 --> 00:32:23,678
for the victims to claim the money back
503
00:32:23,678 --> 00:32:25,578
because it's a legal process.
504
00:32:25,578 --> 00:32:28,638
It takes months or possibly years.
505
00:32:29,418 --> 00:32:31,918
And we don't even know today
506
00:32:31,918 --> 00:32:34,278
which country will be the right place
507
00:32:34,278 --> 00:32:38,898
to basically file for those claimants
508
00:32:38,898 --> 00:32:39,838
to get this money back.
509
00:32:41,118 --> 00:32:42,838
In the U.S., it would cost a lot,
510
00:32:43,018 --> 00:32:44,658
but Taser is not based in the U.S.
511
00:32:44,718 --> 00:32:45,358
They're based somewhere else.
512
00:32:45,418 --> 00:32:48,898
I mean, victims are mostly based in Malaysia,
513
00:32:49,158 --> 00:32:50,018
Indonesia, and so on.
514
00:32:50,058 --> 00:32:52,098
So it's a very difficult situation
515
00:32:52,098 --> 00:32:56,018
for every single victim here.
516
00:32:56,938 --> 00:32:59,058
Can I point something out as well, though?
517
00:33:01,058 --> 00:33:04,118
We don't know who this actor is.
518
00:33:04,278 --> 00:33:09,638
if it's a nation state, if it's some kind of organized crime,
519
00:33:10,018 --> 00:33:13,758
people were joking, making fun of me on X for saying that.
520
00:33:13,818 --> 00:33:14,758
It's like, what's the difference?
521
00:33:17,018 --> 00:33:22,118
They might just want to keep it in Bitcoin and just use it and hold it
522
00:33:22,118 --> 00:33:23,938
maybe for a long period of time.
523
00:33:24,238 --> 00:33:26,998
Maybe they don't care about some.
524
00:33:26,998 --> 00:33:32,298
I could see why you'd want to get it out of Tether or USDC or whatever, right?
525
00:33:32,298 --> 00:33:34,478
because you want to get it out of something that could be frozen.
526
00:33:34,818 --> 00:33:37,498
But depending on, like if this was North Korea,
527
00:33:38,698 --> 00:33:42,238
do they really care about trying to obfuscate where the Bitcoin,
528
00:33:43,818 --> 00:33:46,578
like, you know, or mix the Bitcoin?
529
00:33:46,878 --> 00:33:48,278
They might just want to sit on it.
530
00:33:48,318 --> 00:33:48,758
I don't know.
531
00:33:48,858 --> 00:33:51,438
So we don't know who this actor is.
532
00:33:52,078 --> 00:33:55,318
I'm not sure that the goal might not necessarily be
533
00:33:55,318 --> 00:33:59,358
to just like, you know, coin join and mix all this stuff
534
00:33:59,358 --> 00:34:00,578
and then spend it.
535
00:34:00,578 --> 00:34:08,978
The goal might just be that it could be going right into some treasury somewhere that's just going to be held for decades or longer.
536
00:34:08,978 --> 00:34:12,358
This is a very sophisticated group.
537
00:34:13,798 --> 00:34:19,378
I don't know if the goal is necessarily to obfuscate everything.
538
00:34:19,998 --> 00:34:20,098
Yeah.
539
00:34:20,458 --> 00:34:26,158
Well, it's certainly come out that it's expanded beyond just Asia now, too, right?
540
00:34:26,158 --> 00:34:32,918
We have accounts in Europe of users going out to a media market, the Best Buy of Europe, right?
541
00:34:32,978 --> 00:34:38,038
The big electronics store buying multiple ledgers and they are compromised as well.
542
00:34:38,538 --> 00:34:42,238
Mark, what did you see from these pictures from these ledgers in Europe?
543
00:34:42,638 --> 00:34:51,738
Yes, it was quite interesting actually to see the pictures because the version that was in Europe is actually older than the one that circulated in Malaysia.
544
00:34:51,738 --> 00:35:01,738
which points both to an expanding attack target
545
00:35:01,738 --> 00:35:05,358
but at the same time also it points to a product that is evolving.
546
00:35:06,358 --> 00:35:10,558
This implant, I'm starting to believe,
547
00:35:10,558 --> 00:35:15,898
might be something that some company or someone somewhere
548
00:35:15,898 --> 00:35:21,158
is building and selling to other groups in Malaysia,
549
00:35:21,738 --> 00:35:23,978
maybe in Europe, and we don't know where else.
550
00:35:24,638 --> 00:35:28,918
And those groups received the implants.
551
00:35:29,358 --> 00:35:32,078
They received a manual, and the manual tells them,
552
00:35:32,198 --> 00:35:35,138
you put this on the ledger, you solder here, here, here,
553
00:35:35,378 --> 00:35:37,598
and you close everything like that.
554
00:35:37,738 --> 00:35:40,318
You sell the product, and you're good to go.
555
00:35:40,498 --> 00:35:45,478
And the process has actually become, from earlier version,
556
00:35:45,658 --> 00:35:47,278
much more easy to perform,
557
00:35:47,278 --> 00:35:53,098
which really makes me think this is a product basically.
558
00:35:54,098 --> 00:35:55,258
It is a criminal product.
559
00:35:55,378 --> 00:35:59,438
I don't know if the company selling this is charging per unit,
560
00:35:59,438 --> 00:36:04,778
if they're charging a percentage of the proceeds.
561
00:36:05,618 --> 00:36:08,478
Maybe they even provide money laundering services.
562
00:36:09,278 --> 00:36:11,878
So they provide this with the E-CIMP programs.
563
00:36:11,998 --> 00:36:15,818
They're the one doing the capture of crypto from the seed phrases
564
00:36:15,818 --> 00:36:22,258
and they might be disrupting to the group that actually did the installation part of the process.
565
00:36:22,258 --> 00:36:30,871
So there could be so many options and so many ways this is happening But this product has definitely gone through multiple evolutions
566
00:36:32,811 --> 00:36:38,711
I've re-looked at the very early versions where you just had a few wires connected to the ledger.
567
00:36:38,831 --> 00:36:41,371
And actually, if you look at the board itself, it's labeled.
568
00:36:41,931 --> 00:36:45,451
This is SPI, this is power, and so on.
569
00:36:45,511 --> 00:36:50,871
So I do think that last year, this was still a very basic product.
570
00:36:50,871 --> 00:36:54,651
and very likely they already had some proceeds made
571
00:36:54,651 --> 00:36:59,151
that gave them the option to improve the product
572
00:36:59,151 --> 00:37:02,711
and reach the point it is at right now,
573
00:37:03,251 --> 00:37:04,891
which also makes me believe
574
00:37:04,891 --> 00:37:09,151
that they've got $100 million minus whatever was frozen.
575
00:37:09,711 --> 00:37:11,091
What's going to come next?
576
00:37:12,151 --> 00:37:14,251
And if I could just mention that as well,
577
00:37:14,251 --> 00:37:15,531
that
578
00:37:15,531 --> 00:37:17,651
IoT
579
00:37:17,651 --> 00:37:20,791
chip does have its own
580
00:37:20,791 --> 00:37:22,711
onboard processor
581
00:37:22,711 --> 00:37:24,611
as well. It can run
582
00:37:24,611 --> 00:37:26,931
applications and I know as well
583
00:37:26,931 --> 00:37:28,971
that Nordic
584
00:37:28,971 --> 00:37:30,191
Semi offers
585
00:37:30,191 --> 00:37:32,971
the ability to license through
586
00:37:32,971 --> 00:37:33,391
them
587
00:37:33,391 --> 00:37:36,191
to use their
588
00:37:36,191 --> 00:37:38,291
proprietary SIM
589
00:37:38,291 --> 00:37:39,831
instead of eSIM
590
00:37:39,831 --> 00:37:42,871
but you'd have to be a
591
00:37:42,871 --> 00:37:46,971
company working through them, which maybe is possible for this kind of group to do.
592
00:37:47,371 --> 00:37:53,631
I personally think, I'm curious if Mark agrees, that you could basically get this implant to be
593
00:37:53,631 --> 00:38:01,211
just that chip and on the board. And if you did that, I mean, it's so compact.
594
00:38:02,011 --> 00:38:06,111
You may even be able to use other generations, or I don't know if the other companies make
595
00:38:06,111 --> 00:38:12,451
smaller options, but the right now it has this, this e-sim, it has this microcontroller,
596
00:38:12,671 --> 00:38:17,711
and then it has this IOT chip. And I'm pretty sure you could just get it down to one.
597
00:38:18,591 --> 00:38:24,891
And if that happens, I mean, people were talking about, you know, do a, do a clear case, right?
598
00:38:25,071 --> 00:38:29,671
Weigh the thing. I mean, you could, you could imagine making a battery where this is inside
599
00:38:29,671 --> 00:38:35,171
the battery. You can imagine taping this to the back of the screen where it just looks black.
600
00:38:35,171 --> 00:38:39,091
Like it looks like it's the screen unmodified, but this is inside there.
601
00:38:41,071 --> 00:38:43,951
There's so many different ways that an attacker could do this.
602
00:38:44,091 --> 00:38:45,671
And so I think it's really interesting.
603
00:38:45,791 --> 00:38:51,831
I think Mark kind of exposed this before it was in its like most final optimized form.
604
00:38:52,211 --> 00:38:54,671
And now at least we know what we're on the lookout for.
605
00:38:54,811 --> 00:38:59,631
The flip side of that is I think it's going to get even more compact and more sophisticated.
606
00:38:59,631 --> 00:39:02,551
Like I said, hide it in the screen, hide it in the battery,
607
00:39:02,551 --> 00:39:05,031
maybe even hide it in the plastic
608
00:39:05,031 --> 00:39:06,391
chassis it's so thin
609
00:39:06,391 --> 00:39:09,031
you could do crazy advanced things like that
610
00:39:09,031 --> 00:39:10,111
and
611
00:39:10,111 --> 00:39:12,411
the other thing I want to say is I
612
00:39:12,411 --> 00:39:15,011
am surprised that it only
613
00:39:15,011 --> 00:39:16,071
is affecting ledger
614
00:39:16,071 --> 00:39:18,931
I wonder if we'll see other reports
615
00:39:18,931 --> 00:39:20,671
I mean Trezor is an obvious
616
00:39:20,671 --> 00:39:22,751
you know target Trezor though
617
00:39:22,751 --> 00:39:24,751
ultrasonic welds their chassis
618
00:39:24,751 --> 00:39:26,491
together you would have to
619
00:39:26,491 --> 00:39:29,051
destroy the chassis to get it open
620
00:39:29,051 --> 00:39:30,991
so I think you would need an attacker
621
00:39:30,991 --> 00:39:32,091
who
622
00:39:32,091 --> 00:39:33,871
was just making
623
00:39:33,871 --> 00:39:36,011
Trezor chassis, right, which is
624
00:39:36,011 --> 00:39:37,031
possible.
625
00:39:38,231 --> 00:39:40,011
Ledger is interesting, especially
626
00:39:40,011 --> 00:39:42,191
with the Nano, the older Nano models
627
00:39:42,191 --> 00:39:44,171
because you could just pry them right open without any
628
00:39:44,171 --> 00:39:46,431
kind of, you know, damage
629
00:39:46,431 --> 00:39:48,171
which they've said
630
00:39:48,171 --> 00:39:50,191
is a feature, which I kind of understand, like a customer
631
00:39:50,191 --> 00:39:52,211
can inspect, you know, the internals
632
00:39:52,211 --> 00:39:53,871
but obviously for attackers
633
00:39:53,871 --> 00:39:55,871
an attacker doesn't need to go
634
00:39:55,871 --> 00:39:58,251
deal with injection molding and
635
00:39:58,251 --> 00:40:00,251
you know, all of this stuff to get like
636
00:40:00,251 --> 00:40:02,071
a Ledger device open. I think the
637
00:40:02,091 --> 00:40:03,591
Trezor want to be more sophisticated.
638
00:40:05,711 --> 00:40:07,411
So, yeah, I mean, this is crazy.
639
00:40:07,771 --> 00:40:09,691
Also, the other final thing I want to say is
640
00:40:09,691 --> 00:40:13,031
I wonder if we will see AI-generated images
641
00:40:13,031 -->
00:40:16,971
in the coming days of fake implants
642
00:40:16,971 --> 00:40:19,791
because one of our team members said, I think, today,
643
00:40:19,911 --> 00:40:22,591
if you really want to take down another competitor now,
644
00:40:22,991 --> 00:40:24,351
you just go ask, you know,
645
00:40:24,951 --> 00:40:28,591
ChatGPT to make a picture of whoever's device with an implant
646
00:40:28,591 --> 00:40:30,531
and now everyone will panic, right?
647
00:40:30,531 --> 00:40:35,391
So I think there's going to be a bit of a fog of war kind of thing in the coming days as well.
648
00:40:35,931 --> 00:40:37,851
I don't think you can trust the photo.
649
00:40:37,991 --> 00:40:45,971
I think you need to trust that someone mails it, you know, to someone like Mark to look at and have like a reputable person to actually verify that this happened.
650
00:40:46,371 --> 00:40:52,951
If someone goes and posts an image of Trezor tomorrow, I'm probably not going to believe it unless it came from someone that I trust.
651
00:40:53,011 --> 00:40:57,931
So this is it's going to be, I think, a crazy few days and weeks and so on.
652
00:40:57,931 --> 00:41:01,511
How many apps does it take to run your life on Bitcoin?
653
00:41:01,851 --> 00:41:05,011
One to buy, one to send, one for lightning and another for loans.
654
00:41:05,831 --> 00:41:07,391
That's a lot of friction.
655
00:41:07,931 --> 00:41:09,231
Rhino replaces all of them.
656
00:41:09,571 --> 00:41:13,671
Stay to the end and I'll tell you how Rhino can set you up with $21 in free Bitcoin.
657
00:41:13,991 --> 00:41:18,851
Rhino is a next generation Bitcoin app built for actually living on Bitcoin, not just holding it.
658
00:41:19,331 --> 00:41:20,931
Buy Bitcoin with zero fees.
659
00:41:21,311 --> 00:41:23,071
Borrow against your stack without selling it.
660
00:41:23,071 --> 00:41:27,371
Move money globally on Bitcoin rails in minutes rather than a wire that takes days.
661
00:41:27,931 --> 00:41:32,311
Set a DCA plan so your stack keeps growing, whether the market is up, down,
662
00:41:32,431 --> 00:41:33,931
or you're out touching grass.
663
00:41:33,931 --> 00:41:37,811
Bill pay, cards and Bitcoin retirement accounts are all on the way.
664
00:41:38,151 --> 00:41:39,991
And your Bitcoin is never worth.
665
00:41:39,997 --> 00:41:44,197
be hypothecated, never lent out and never touched by anyone other than you.
666
00:41:44,517 --> 00:41:49,657
Biometric protection, 24-7 fraud monitoring, and you can move to full self-custody whenever
667
00:41:49,657 --> 00:41:55,597
you're ready. It's your Bitcoin always. Download Rhino for iOS or Android at the link below and
668
00:41:55,597 --> 00:42:02,077
use code Bitcoin News. Deposit your first $100 USD and Rhino drops you $21 in free Bitcoin
669
00:42:02,077 --> 00:42:06,017
straight into your wallet. At Rhino, you get the most Bitcoin for your dollar.
670
00:42:06,017 --> 00:42:13,557
yeah and mark over to you why do you think ledger was the first one to have to deal with this they're
671
00:42:13,557 --> 00:42:18,377
definitely the biggest but is there anything in particular yeah i do think there's very much
672
00:42:18,377 --> 00:42:26,057
an evaluation of the cost versus the potential return and uh in this case ledger is the largest
673
00:42:26,057 --> 00:42:33,797
uh it also turned out to be fairly easy to add the implants and because everyone just
674
00:42:33,797 --> 00:42:35,637
used to ledgers, I don't even check
675
00:42:35,637 --> 00:42:37,877
inside the ledgers, but it
676
00:42:37,877 --> 00:42:39,177
being fairly easy to do,
677
00:42:39,937 --> 00:42:41,757
which resulted in this
678
00:42:41,757 --> 00:42:43,877
situation. I've
679
00:42:43,877 --> 00:42:45,837
seen a few pictures of
680
00:42:45,837 --> 00:42:47,597
AI-generated implants too,
681
00:42:48,057 --> 00:42:49,997
some easier to recognize
682
00:42:49,997 --> 00:42:51,697
than others, but typically
683
00:42:51,697 --> 00:42:53,937
AI cannot reliably
684
00:42:53,937 --> 00:42:55,177
generate
685
00:42:55,177 --> 00:42:57,637
real-looking circuit
686
00:42:57,637 --> 00:42:58,657
bonds. So at this point,
687
00:42:59,197 --> 00:43:01,417
we're still able to make the difference.
688
00:43:02,897 --> 00:43:03,777
When the
689
00:43:03,777 --> 00:43:06,817
I think it was a stack was posted by someone.
690
00:43:07,937 --> 00:43:14,237
I first suspected AI because that was the most likely case,
691
00:43:14,377 --> 00:43:16,937
but I really closely inspected the pictures,
692
00:43:17,077 --> 00:43:19,897
looked at how it was installed, how it was set up,
693
00:43:20,637 --> 00:43:22,497
and determined that it was real.
694
00:43:23,577 --> 00:43:26,857
But yes, there's really a question of detecting,
695
00:43:27,037 --> 00:43:29,137
basically recognizing AI in those pictures.
696
00:43:30,937 --> 00:43:32,937
But there's also another issue that I think
697
00:43:32,937 --> 00:43:38,257
that we're going to be seeing is if I was to be developing
698
00:43:38,257 --> 00:43:40,417
and creating those implants,
699
00:43:40,537 --> 00:43:42,337
and if I had maybe even $5 million,
700
00:43:42,857 --> 00:43:47,517
I could build something that would look exactly like this.
701
00:43:47,957 --> 00:43:49,377
This is a regular board.
702
00:43:49,897 --> 00:43:52,537
You couldn't make a difference from this.
703
00:43:53,737 --> 00:43:56,637
And even myself wouldn't be able to make a difference
704
00:43:56,637 --> 00:43:57,277
just by looking at it.
705
00:43:57,317 --> 00:43:59,117
I would have to bring inside an X-ray machine
706
00:43:59,117 --> 00:44:01,817
or this kind of things to be able to differentiate
707
00:44:01,817 --> 00:44:05,057
an implanted device from a real device,
708
00:44:05,897 --> 00:44:09,837
which means that right now it's easy.
709
00:44:09,937 --> 00:44:11,217
You can open your device, look inside.
710
00:44:11,297 --> 00:44:12,997
You will see if it's an implanted device or not.
711
00:44:13,897 --> 00:44:16,057
I don't know how long it will take
712
00:44:16,057 --> 00:44:17,897
for these kind of attackers to switch
713
00:44:17,897 --> 00:44:22,497
to missiles that are much more stealthy.
714
00:44:24,197 --> 00:44:28,677
I do guess there might be some other kind of attackers,
715
00:44:28,677 --> 00:44:31,657
like for example targeting the U.S. military and so on,
716
00:44:31,817 --> 00:44:37,497
on the country versus country size that may already be using this kind of very advanced
717
00:44:39,497 --> 00:44:46,137
basically hidden shapes and that it takes exploitation methods that cost more to develop.
718
00:44:48,057 --> 00:44:53,977
But even with a fairly low cost per device today, the way the Ledger for the company is built
719
00:44:54,937 --> 00:44:59,817
has a number of flows that could be improved, of course, and I will try to get in touch with
720
00:44:59,817 --> 00:45:00,777
with the jar about that.
721
00:45:02,397 --> 00:45:03,917
But it would be possible
722
00:45:03,917 --> 00:45:05,897
to make a board that is
723
00:45:05,897 --> 00:45:08,157
I mean, you kind of distinguish it.
724
00:45:08,177 --> 00:45:09,637
You look at it, it looks the same.
725
00:45:09,657 --> 00:45:10,417
It has the same component.
726
00:45:11,757 --> 00:45:13,857
You cannot find anything because
727
00:45:13,857 --> 00:45:17,357
I mean, you would need an explanation
728
00:45:17,357 --> 00:45:19,417
to basically what's hidden in it.
729
00:45:20,977 --> 00:45:22,237
But it would be possible
730
00:45:22,237 --> 00:45:22,917
to create a method
731
00:45:22,917 --> 00:45:25,717
that would go that far.
732
00:45:27,677 --> 00:45:29,237
It's only going to get crazier.
733
00:45:29,237 --> 00:45:30,677
It's only going to get crazier, right?
734
00:45:30,917 --> 00:45:36,157
So, and I think that leads us naturally to where does the industry go from here?
735
00:45:36,337 --> 00:45:40,077
I think, Zach, you put out a good post about how you see it bifurcating.
736
00:45:40,417 --> 00:45:41,297
Tell us about that.
737
00:45:41,597 --> 00:45:43,437
Yeah, I see kind of two directions.
738
00:45:43,837 --> 00:45:49,317
One direction is you go like generic hardware or DIY hardware.
739
00:45:49,917 --> 00:45:51,397
You build it yourself.
740
00:45:52,757 --> 00:45:58,897
You lose a lot of those security properties, but you know that you built it yourself, right?
741
00:45:58,897 --> 00:46:01,997
and you are responsible for storing those seed words.
742
00:46:02,097 --> 00:46:05,137
I mean, SeedSigner is, I think, a fantastic project
743
00:46:05,137 --> 00:46:08,077
and is a good example.
744
00:46:08,797 --> 00:46:10,477
People grabbing old laptops,
745
00:46:10,937 --> 00:46:13,557
people making jokes to get a laptop from before Bitcoin
746
00:46:13,557 --> 00:46:15,457
was even a thing.
747
00:46:15,637 --> 00:46:17,717
I mean, I think there's pros and cons.
748
00:46:17,837 --> 00:46:19,617
I think if you're a sophisticated person,
749
00:46:19,617 --> 00:46:21,117
if you're able to do that,
750
00:46:21,237 --> 00:46:24,597
then you probably can figure out how to do it in a secure way.
751
00:46:24,597 --> 00:46:32,577
I think the other side of it is you go hardcore security features, right?
752
00:46:32,657 --> 00:46:36,537
You go, you take what we have like in Passport Prime,
753
00:46:36,677 --> 00:46:40,897
we have a couple of these tamper switches that actuate,
754
00:46:41,477 --> 00:46:46,157
you know, when you, like, you know, you take our board,
755
00:46:46,277 --> 00:46:59,210
you take the screen you know you push it together it like physically depresses a switch You just go much more hardcore than that You take it from like the payment card industry you do tamper meshes you take it from the HSM industry
756
00:46:59,210 --> 00:47:01,350
where they hardware security modules
757
00:47:01,350 --> 00:47:03,750
where they will encase the entire thing
758
00:47:03,750 --> 00:47:06,710
in like an epoxy, you call that a potting compound.
759
00:47:07,450 --> 00:47:09,530
You run, I figured out also,
760
00:47:09,770 --> 00:47:11,490
like you would probably think this is cool, Mark.
761
00:47:11,890 --> 00:47:13,050
I figured out over the weekend,
762
00:47:13,050 --> 00:47:16,390
like if we're specking out our display from our provider,
763
00:47:16,390 --> 00:47:20,090
we can run our tamper circuit through the display cable.
764
00:47:21,050 --> 00:47:24,190
You can make it where if anyone was to remove that display,
765
00:47:24,190 --> 00:47:25,410
that would also trigger the tamper.
766
00:47:25,470 --> 00:47:26,430
You can run it through everything.
767
00:47:26,550 --> 00:47:27,330
You run it through the board.
768
00:47:27,630 --> 00:47:29,090
You put a cage over the parts.
769
00:47:29,450 --> 00:47:31,050
You fill it with epoxy.
770
00:47:31,790 --> 00:47:34,410
You put light sensors if someone opens it.
771
00:47:34,470 --> 00:47:35,810
You have the temperature sensors.
772
00:47:36,510 --> 00:47:38,190
You can even do impedance type stuff
773
00:47:38,190 --> 00:47:40,270
where you can detect if someone is plugging in,
774
00:47:40,530 --> 00:47:41,870
tapping those lines, right?
775
00:47:42,090 --> 00:47:45,870
Trying to power other types of parts off of those lines.
776
00:47:45,870 --> 00:47:48,050
so you just go as hardcore as possible
777
00:47:48,050 --> 00:47:49,450
and I feel like there's
778
00:47:49,450 --> 00:47:50,990
within that world
779
00:47:50,990 --> 00:47:53,030
there's also the world of like
780
00:47:53,030 --> 00:47:54,350
you still do open source
781
00:47:54,350 --> 00:47:55,490
and you make it auditable
782
00:47:55,490 --> 00:47:58,130
which is our approach
783
00:47:58,130 --> 00:47:59,650
where like you do all this
784
00:47:59,650 --> 00:48:01,530
but it's still you know open source
785
00:48:01,530 --> 00:48:03,730
versus or you go the opposite
786
00:48:03,730 --> 00:48:04,810
you probably kind of continue
787
00:48:04,810 --> 00:48:05,570
the ledger approach
788
00:48:05,570 --> 00:48:06,670
where it's all proprietary
789
00:48:06,670 --> 00:48:08,550
right you don't talk about
790
00:48:08,550 --> 00:48:09,930
how all that stuff works
791
00:48:09,930 --> 00:48:11,630
and you go down that road
792
00:48:11,630 --> 00:48:13,430
kind of like the traditional smart card
793
00:48:13,430 --> 00:48:14,950
or HSM industry that we have
794
00:48:14,950 --> 00:48:23,290
I mean, we're going to go down that road of making it virtually impossible to get into this thing without triggering a tamper.
795
00:48:23,430 --> 00:48:24,650
But we have to be careful.
796
00:48:24,770 --> 00:48:25,850
I mean, we need you.
797
00:48:26,010 --> 00:48:27,750
It requires internal batteries.
798
00:48:28,470 --> 00:48:31,770
So the question is, can you still make it user accessible?
799
00:48:32,030 --> 00:48:36,770
Can you enable a battery replacement or does you just throw it out after five years?
800
00:48:36,770 --> 00:48:39,670
Like there's a lot of those questions that we're trying to figure out.
801
00:48:39,670 --> 00:48:47,630
And then the other thing I would say is that you have to figure out how to make sure that the customer doesn't get social engineered, which I think is even harder.
802
00:48:47,910 --> 00:48:50,710
Because like I said, you can do everything I just described.
803
00:48:51,230 --> 00:49:02,110
But if they get a device that looks like our device and you turn it on and it says, thank you, time to set up your passport by foundation, scan the QR code to download our app.
804
00:49:02,430 --> 00:49:04,290
And it takes you to a different app.
805
00:49:04,290 --> 00:49:11,190
Well, now you've bypassed the entire supply chain validation process and you can be completely screwed.
806
00:49:11,190 --> 00:49:19,690
And so we were talking today, like, do we require that you buy a passport from our mobile app so that like you have to have the mobile app?
807
00:49:19,830 --> 00:49:22,750
But even then you could probably turn on the thing.
808
00:49:23,230 --> 00:49:34,210
And I mean, there's some level of education here to the customers where you you have to be told, like, it's not enough to just power on this device and follow the instructions.
809
00:49:34,210 --> 00:49:36,110
that came in a box that looks like it's from us.
810
00:49:36,150 --> 00:49:38,050
So all those things I described,
811
00:49:38,150 --> 00:49:39,690
the industry has to be working on.
812
00:49:39,790 --> 00:49:40,830
And I really hope they do.
813
00:49:41,290 --> 00:49:45,790
If we had even a 10% R&D budget of what Ledger had,
814
00:49:45,990 --> 00:49:49,230
I think we would have already done a lot more of this.
815
00:49:49,750 --> 00:49:50,890
Yeah, and Mark, where do you see it going?
816
00:49:51,430 --> 00:49:52,990
Yeah, I mean, one of the solutions I see
817
00:49:52,990 --> 00:49:54,030
that is very promising
818
00:49:54,030 --> 00:49:58,570
is basically multi-party computation.
819
00:49:59,710 --> 00:50:02,250
I'm trying to see if we could,
820
00:50:02,250 --> 00:50:05,110
I mean, I'm drafting right now, actually,
821
00:50:05,950 --> 00:50:08,630
a multi-vendor, multi-party communication,
822
00:50:08,630 --> 00:50:13,170
which would provide a protocol that would allow multiple hardware wallets
823
00:50:13,170 --> 00:50:18,430
of different vendors to act as different parts of a single,
824
00:50:18,750 --> 00:50:21,390
basically, public key.
825
00:50:22,090 --> 00:50:26,890
So you have your Bitcoin address, but for actually using it,
826
00:50:26,930 --> 00:50:30,590
you need at least two different wallets that you own.
827
00:50:32,250 --> 00:50:37,190
And this means that even if, for example, one is compromised, it cannot do anything.
828
00:50:37,990 --> 00:50:39,410
There are different ways to configure that.
829
00:50:39,570 --> 00:50:43,190
You could have five devices and require four of the devices.
830
00:50:44,650 --> 00:50:52,250
Multiparty computation is kind of, I like to call this black magic of ECTSA,
831
00:50:52,250 --> 00:50:56,510
but it's actually compatible with other protocols, including post-content protocols.
832
00:50:56,510 --> 00:51:06,990
So this is, for me, one of the most promising options in the future in the sense that even
833
00:51:06,990 --> 00:51:10,910
customers who, for example, only want to have one hardware wallet can also have their mobile
834
00:51:10,910 --> 00:51:11,630
phone participating.
835
00:51:13,050 --> 00:51:17,090
And the third part could be held by a vendor.
836
00:51:17,990 --> 00:51:23,970
This way, you have, for example, SMS authentication with a vendor outside of your home in case
837
00:51:23,970 --> 00:51:25,750
of something, you lose your actual wallet.
838
00:51:25,750 --> 00:51:28,270
you have your physical wallet, you have your mobile phone
839
00:51:28,270 --> 00:51:30,650
and with just what you have on hand
840
00:51:30,650 --> 00:51:32,450
you can send crypto
841
00:51:32,450 --> 00:51:34,130
or do everything you need to do
842
00:51:34,130 --> 00:51:35,890
and if you lose Ether
843
00:51:35,890 --> 00:51:38,410
you can still recover
844
00:51:38,410 --> 00:51:40,790
by using that third party
845
00:51:40,790 --> 00:51:43,630
Ian, if I could add to that
846
00:51:43,630 --> 00:51:45,590
just from the Bitcoin perspective
847
00:51:45,590 --> 00:51:47,950
because a lot of Bitcoiners think multi-sig
848
00:51:47,950 --> 00:51:50,470
but traditional multi-sig
849
00:51:50,470 --> 00:51:51,910
has I think a bunch of problems
850
00:51:51,910 --> 00:51:53,090
a big one being that
851
00:51:53,090 --> 00:51:56,230
you can't swap signers
852
00:51:56,230 --> 00:51:58,270
without doing a full rotation
853
00:51:58,270 --> 00:51:59,550
of a key rotation
854
00:51:59,550 --> 00:52:11,863
and a sweep of all the funds And so you know MPC right in this case not only is it compatible with any chain right more than just Bitcoin but you can you could adopt you could have a quorum
855
00:52:11,863 --> 00:52:18,363
of signers, you can make changes on the fly without having to move the funds. Frost on Bitcoin,
856
00:52:18,743 --> 00:52:24,483
you know, can do this as well as an example of when Bitcoiners say Frost multisig,
857
00:52:24,483 --> 00:52:29,903
it's really less like multi-sig and more like multi-party computation so i think it's very
858
00:52:29,903 --> 00:52:38,343
promising my concern mark though is like we we use our devices now um for more than just storing
859
00:52:38,343 --> 00:52:46,283
bitcoin we do fido security keys and six digit 2fa codes and file storage and other stuff and so
860
00:52:46,283 --> 00:52:50,683
that's where i struggle with it where i almost feel maybe a two-tier system where
861
00:52:50,683 --> 00:52:53,523
that you can still have like a primary device, right,
862
00:52:53,563 --> 00:52:56,143
that has utility and does all this stuff.
863
00:52:56,343 --> 00:52:58,763
But maybe when you're storing your money,
864
00:52:59,063 --> 00:53:01,743
you have that protected by this, you know,
865
00:53:02,303 --> 00:53:04,523
MPC or multi-sig type scheme.
866
00:53:04,683 --> 00:53:07,203
Because, yeah, if I lose my 2FA codes,
867
00:53:07,303 --> 00:53:08,483
look, it's not the end of the world.
868
00:53:08,903 --> 00:53:11,563
But if I lose my Bitcoin, it kind of is.
869
00:53:12,623 --> 00:53:12,783
Yeah.
870
00:53:13,323 --> 00:53:16,223
And that's probably the interesting point is
871
00:53:16,223 --> 00:53:20,663
the way I'd see it is you can set up MPC
872
00:53:20,663 --> 00:53:22,503
with three hardware wallets,
873
00:53:23,063 --> 00:53:25,423
or you can set it up just with your mobile phone.
874
00:53:26,003 --> 00:53:27,723
And depending on how much you're storing,
875
00:53:27,863 --> 00:53:29,283
you can decide what is best for you.
876
00:53:29,503 --> 00:53:32,903
So if you're storing, I don't know, like $100,
877
00:53:33,503 --> 00:53:37,563
you might not want to have to get out all your devices
878
00:53:37,563 --> 00:53:39,303
each time you want to sign something.
879
00:53:39,423 --> 00:53:40,643
But if you're storing larger amounts,
880
00:53:40,723 --> 00:53:43,043
you will want to invest maybe not three,
881
00:53:43,043 --> 00:53:45,823
but five or more machines,
882
00:53:45,943 --> 00:53:49,103
and you store one inside a bank vault somewhere, and so on.
883
00:53:49,103 --> 00:53:55,343
So there are different strategies you can apply based on your situation, basically.
884
00:53:56,003 --> 00:54:07,603
And when it comes to self-custody itself, I think we all believe that it's a fundamental part of Bitcoin and that it's important for Bitcoiners to know how to do it, use it, believe in it.
885
00:54:07,603 --> 00:54:16,743
But when it comes to getting new people on board, I mean, everything you guys just said sounds extremely complicated, technical and scary to the average person.
886
00:54:17,003 --> 00:54:18,803
Right. So what's the new pitch?
887
00:54:19,103 --> 00:54:21,283
In the face of all the attacks that we've just seen,
888
00:54:21,483 --> 00:54:24,183
how do we onboard the new people into self-custody
889
00:54:24,183 --> 00:54:26,663
after this summer of nightmares?
890
00:54:27,503 --> 00:54:27,803
Zach.
891
00:54:29,863 --> 00:54:31,263
I have all the answers.
892
00:54:33,763 --> 00:54:35,243
No, I mean, I think,
893
00:54:36,003 --> 00:54:38,943
one is I think it still all comes down
894
00:54:38,943 --> 00:54:40,243
to the actual user experience.
895
00:54:40,283 --> 00:54:41,723
Like everything Mark just described
896
00:54:41,723 --> 00:54:43,563
could be made very, very easy
897
00:54:43,563 --> 00:54:45,343
with the mobile app, with the device.
898
00:54:45,903 --> 00:54:48,623
You're just tapping on two different places,
899
00:54:48,623 --> 00:54:50,003
to approve something.
900
00:54:50,763 --> 00:54:52,963
I don't think it needs to be complicated.
901
00:54:53,243 --> 00:54:56,123
I think the problem is that so far,
902
00:54:56,223 --> 00:54:58,083
most of the leaders of this industry
903
00:54:58,083 --> 00:55:00,543
have made things way too complicated
904
00:55:00,543 --> 00:55:06,783
and have either said that you need to keep it
905
00:55:06,783 --> 00:55:09,143
intentionally difficult to use,
906
00:55:09,663 --> 00:55:11,823
which is, I would say, the cold card approach,
907
00:55:12,363 --> 00:55:14,603
make it very difficult and unapproachable,
908
00:55:15,023 --> 00:55:17,403
or you try to make it so approachable
909
00:55:17,403 --> 00:55:20,083
that it gets creepy, like the Ledger Recover stuff
910
00:55:20,083 --> 00:55:22,363
where you have to use like your driver's license
911
00:55:22,363 --> 00:55:25,703
and what on, KYC and three custodians.
912
00:55:25,943 --> 00:55:28,523
And so I think you could do it as Mark described.
913
00:55:28,643 --> 00:55:30,583
It's just, it's a serious UX challenge
914
00:55:30,583 --> 00:55:33,643
that I think is totally possible to achieve.
915
00:55:34,863 --> 00:55:37,363
Yeah, I think the most important is
916
00:55:37,363 --> 00:55:39,203
you don't have one size fits all.
917
00:55:39,343 --> 00:55:41,463
You need to have different solutions for different people.
918
00:55:42,463 --> 00:55:44,083
For sure. Yeah, for sure.
919
00:55:44,083 --> 00:55:48,423
I would also say though that this,
920
00:55:48,543 --> 00:55:52,083
our hardware wallet industry needs to seriously professionalize.
921
00:55:53,563 --> 00:55:55,923
It's such a,
922
00:55:57,883 --> 00:56:02,603
all the security practices that are employed by the industries that I have
923
00:56:02,603 --> 00:56:07,283
historically criticized, the HSM world, the smart card world,
924
00:56:07,803 --> 00:56:10,723
you know, all very legacy proprietary security stuff.
925
00:56:10,723 --> 00:56:14,163
you have eight different security certifications.
926
00:56:15,263 --> 00:56:17,723
You know, everything is ultra proprietary and secret.
927
00:56:18,063
--> 00:56:20,323
You have, like I mentioned, all your tamper
928
00:56:20,323 --> 00:56:24,203
and your tamper meshes and your potting material
929
00:56:24,203 --> 00:56:25,043
and all this stuff.
930
00:56:27,383 --> 00:56:29,363
Just because like I'm critical
931
00:56:29,363 --> 00:56:32,103
of the proprietary nature of that world
932
00:56:32,103 --> 00:56:37,063
does not mean that you shouldn't go do all that stuff.
933
00:56:37,623 --> 00:56:39,003
You could still do it all
934
00:56:39,003 --> 00:56:42,403
and have your hardware and software be open source, right?
935
00:56:42,403 --> 00:56:44,383
You can get all the security certifications,
936
00:56:44,963 --> 00:56:46,443
some of which are very stringent,
937
00:56:47,123 --> 00:56:48,743
like FIPS certification,
938
00:56:49,323 --> 00:56:54,163
which if you have HSMs or security hardware in the US
939
00:56:54,163 --> 00:56:56,163
that's purchased by like the military or government,
940
00:56:56,783 --> 00:56:58,243
it's FIPS certified.
941
00:56:58,403 --> 00:57:01,623
Yubikey sells a more expensive option
942
00:57:01,623 --> 00:57:03,123
that has a FIPS certification.
943
00:57:03,283 --> 00:57:04,003
There's different levels.
944
00:57:04,143 --> 00:57:06,003
There's level one, level two, level three.
945
00:57:06,323 --> 00:57:07,723
There's no hardware wallets that have that.
946
00:57:08,103 --> 00:57:08,983
Ledger doesn't have that.
947
00:57:09,003 --> 00:57:10,023
No one has it.
948
00:57:11,783 --> 00:57:23,297
There all the and all the other stuff I mentioned like the tamper and the potting and the meshes and everything like that Foundation going to go after all that stuff I mean there has to be a it going to be expensive It going to take time
949
00:57:24,517 --> 00:57:29,037
But there has to be like the companies in this space have to have to be professional.
950
00:57:29,037 --> 00:57:39,297
You know, you can see from this ledger stuff, all the high tech practices that are employed in a credit card terminal, right?
951
00:57:39,797 --> 00:57:41,937
There's no physical protection.
952
00:57:42,477 --> 00:57:45,697
You can crack open a ledger and you can keep using it.
953
00:57:46,377 --> 00:57:50,437
And I think the industry itself has to professionalize.
954
00:57:50,557 --> 00:57:54,097
Also, someone asked me something on X today or yesterday,
955
00:57:54,197 --> 00:57:55,677
which I thought was very provocative.
956
00:57:55,677 --> 00:58:01,297
They asked us, are we a crypto company or are we a security company?
957
00:58:02,537 --> 00:58:05,577
And I answered, we're a security company.
958
00:58:06,437 --> 00:58:12,017
And I feel like too often everyone in the hardware wallet world,
959
00:58:12,017 --> 00:58:15,157
all the companies are thinking that they're in the Bitcoin or crypto space.
960
00:58:15,817 --> 00:58:21,537
But really, you could make an ultra secure device that has all these features I mentioned
961
00:58:21,537 --> 00:58:23,497
and have nothing
to do with crypto.
962
00:58:24,137 --> 00:58:28,917
You could also make a Bitcoin-only air-gapped hardware wallet
963
00:58:28,917 --> 00:58:32,297
that doesn't even have any entropy, right?
964
00:58:32,377 --> 00:58:35,257
And so we really need to think about how we define ourselves.
965
00:58:35,577 --> 00:58:38,457
And I think it's time to maybe define ourselves more as like,
966
00:58:38,457 --> 00:58:43,617
we're making security hardware as opposed to, oh yeah, we make like crypto, you know,
967
00:58:43,657 --> 00:58:48,297
hardware wallets and psychologically and the way that you build that business and,
968
00:58:48,297 --> 00:58:53,717
and build these products and, and, and, and so on is kind of all downstream of that.
969
00:58:53,837 --> 00:58:55,477
And so I think it's, it's very important.
970
00:58:56,217 --> 00:59:03,837
Yes. And in closing here, Mark, being in Bitcoin for a long time is a rollercoaster
971
00:59:03,837 --> 00:59:06,157
ups and downs, the good and the bad.
972
00:59:07,117 --> 00:59:11,217
This is just the latest punch in the face that this industry has taken.
973
00:59:12,057 --> 00:59:13,037
What do you think?
974
00:59:13,197 --> 00:59:15,177
Looking forward, are you still optimistic?
975
00:59:15,417 --> 00:59:19,297
Do you still believe in the vision of Bitcoin and what it can provide for the world?
976
00:59:19,917 --> 00:59:21,797
Yes, I still believe in Bitcoin.
977
00:59:21,797 --> 00:59:27,517
And one thing I do think that will be extremely important is decentralization.
978
00:59:28,257 --> 00:59:33,197
Because the more people use different options, different ways to protect their Bitcoins,
979
00:59:33,197 --> 00:59:36,577
the more difficult it would be for an attacker
980
00:59:36,577 --> 00:59:39,837
to basically invest a lot on one attack
981
00:59:39,837 --> 00:59:43,017
and make a big amount.
982
00:59:43,977 --> 00:59:46,437
So as Zach says,
983
00:59:46,577 --> 00:59:48,697
it's people that may be better using
984
00:59:48,697 --> 00:59:53,497
more compliant traditional security systems
985
00:59:53,497 --> 00:59:55,997
where some people may do better with
986
00:59:55,997 --> 01:00:00,597
do-it-yourself when they've built from Parzivit.
987
01:00:00,597 --> 01:00:03,957
and it's actually the best way
988
01:00:03,957 --> 01:00:06,757
that everyone has their option.
989
01:00:06,877 --> 01:00:08,497
Of course, each one needs to be secure,
990
01:00:08,757 --> 01:00:10,337
tested and so on.
991
01:00:10,357 --> 01:00:11,897
So there's still a lot of work to be done.
992
01:00:13,517 --> 01:00:15,697
It's kind of weird that we don't have
993
01:00:15,697 --> 01:00:17,897
any FIB certified wallet at this point, actually.
994
01:00:19,117 --> 01:00:21,717
I know the certification is not that expensive to get,
995
01:00:21,937 --> 01:00:23,077
especially for a wallet company.
996
01:00:23,077 --> 01:00:26,217
I would expect more companies to have that.
997
01:00:26,217 --> 01:00:29,737
but yeah basically
998
01:00:29,737 --> 01:00:31,317
the main
999
01:00:31,317 --> 01:00:33,957
the main strength of Bitcoin
1000
01:00:33,957 --> 01:00:36,077
is to not be in one
1001
01:00:36,077 --> 01:00:37,657
single place because
1002
01:00:37,657 --> 01:00:40,397
it means that you cannot take down Bitcoin
1003
01:00:40,397 --> 01:00:42,137
if you wanted to
1004
01:00:42,137 --> 01:00:44,137
stop the network you would have to basically
1005
01:00:44,137 --> 01:00:46,217
wipe every single running node
1006
01:00:46,217 --> 01:00:47,677
out of existence which is not
1007
01:00:47,677 --> 01:00:49,417
something that will ever happen
1008
01:00:49,417 --> 01:00:51,937
and the same applies to custody if everyone just
1009
01:00:51,937 --> 01:00:54,197
stores their Bitcoin the same
1010
01:00:54,197 --> 01:00:56,037
way it means that for an
1011
01:00:56,037 --> 01:01:00,797
attacker you just target this specific way and you get a lot of money out of it so
1012
01:01:00,797 --> 01:01:09,477
this one tradition this one station of custody is I do believe the best
1013
01:01:09,477 --> 01:01:15,717
direction to make this kind of attacks less profitable yeah in the words of
1014
01:01:15,717 --> 01:01:20,937
Jeff booth as long as Bitcoin stays decentralized and secure its future is
1015
01:01:20,937 --> 01:01:26,957
is bright. Well, I feel extremely honored to have you two on this same conversation with
1016
01:01:26,957 --> 01:01:33,637
us here tonight. I think it's very important for the community to hear the takeaways from
1017
01:01:33,637 --> 01:01:40,517
this brutal attack. But we will find our way through as Bitcoin tends to do. Thank you
1018
01:01:40,517 --> 01:01:42,177
so much for joining us today.
1019
01:01:42,177 --> 01:01:43,177
Thanks, Mark.
1020
01:01:43,177 --> 01:01:44,177
Thank you very much.
1021
01:01:44,177 --> 01:01:45,177
Thanks, Mark.
1022
01:01:45,177 --> 01:01:46,177
Cheers. Bye-bye.
1023
01:01:46,177 --> 01:01:50,817
Most Bitcoiners start the same way. You buy some Bitcoin on an exchange, then you get
1024
01:01:50,817 --> 01:01:55,057
a bit more serious. Decide it's time for a hardware wallet and take self-custody of your
1025
01:01:55,057 --> 01:01:59,417
Bitcoin. Eventually, you start thinking about running your own node or maybe even getting
1026
01:01:59,417 --> 01:02:03,917
into home mining. But what if you could build out your entire Bitcoin stack in one place?
1027
01:02:04,337 --> 01:02:09,757
That's where our sponsor, Solo Satoshi, comes in. Pick your hardware a la carte or grab one of
1028
01:02:09,757 --> 01:02:14,797
their new bundles and get everything you need in one go. They've got hardware wallets, Bitcoin nodes,
1029
01:02:15,057 --> 01:02:19,737
home servers, solo miners, and even dice and entropy tablets if you want to go all the way
1030
01:02:19,737 --> 01:02:22,137
down the rabbit hole and generate your own entropy.
1031
01:02:22,137 --> 01:02:24,357
Having sold over 50,000 devices,
1032
01:02:24,357 --> 01:02:27,137
Solo Satoshi carries the brands you know and love,
1033
01:02:27,137 --> 01:02:29,977
including Bidex, Blockstream, Start9, Seedline,
1034
01:02:29,982 --> 01:02:31,982
designer, foundation, and Kanan.
1035
01:02:31,982 --> 01:02:36,002
Solo Satoshi is building the one-stop shop for the sovereign bitcoiner.
1036
01:02:36,002 --> 01:02:40,162
Whether you're looking for the latest cutting-edge home miner, upgrading your self-custody setup,
1037
01:02:40,162 --> 01:02:44,202
or just putting together a starter bundle for your mom, they've got you covered.
1038
01:02:44,202 --> 01:02:48,562
Find the link to solosatoshi.com in the description below.
1039
01:02:48,562 --> 01:02:51,322
Solo Satoshi, your trusted source for Bitcoin hardware.
1040
01:02:59,982 --> 01:03:00,922
Thank you.