رونوشت 🔗
WEBVTT
1
00:00:05.759 --> 00:00:12.160
Large scale destructive attacks on big machinery. It's not something
2
00:00:12.160 --> 00:00:14.359
that I would consider a credible attack.
3
00:00:19.679 --> 00:00:24.199
Welcome, Everyone's the Industrial Security Podcast. My name is Nate Nelson.
4
00:00:24.679 --> 00:00:28.120
I'm here with Andrew Ginter, the vice president of Industrial
5
00:00:28.160 --> 00:00:32.119
Security at Waterfall Security Solutions, who's going to introduce the
6
00:00:32.159 --> 00:00:35.799
subject and guest of our show today. Andrew, how are you.
7
00:00:36.759 --> 00:00:39.359
I'm very well, Thank you, Nate. Our guest today is
8
00:00:39.520 --> 00:00:43.399
Kenneth Tittelstad. He is the chief commercial Officer at Omni
9
00:00:44.320 --> 00:00:48.560
and he's also the chair of the Norwegian International Electrotechnical
10
00:00:48.640 --> 00:00:52.840
Committee Subgroup working on sixty two four four to three.
11
00:00:52.920 --> 00:00:56.119
So this is the Norwegian delegation to the IEC that
12
00:00:56.200 --> 00:00:59.359
produces the widely used IEC six two four four three standard.
13
00:01:00.479 --> 00:01:04.519
We're going to be talking about credible threats, what should
14
00:01:04.519 --> 00:01:07.560
we be planning for security wise? And by the way
15
00:01:08.319 --> 00:01:11.560
I happened, I had an opportunity to be in Norway
16
00:01:12.079 --> 00:01:15.879
and I visited Kenneth at the OMNI head office where
17
00:01:15.879 --> 00:01:19.200
they have a lovely recording studio. So we recorded this
18
00:01:19.280 --> 00:01:22.359
face to face in their studio in their head office.
19
00:01:23.000 --> 00:01:26.840
Then let's get right into your conversation with Kenneth.
20
00:01:29.879 --> 00:01:33.040
Hello, Kenneth, and welcome to the podcast. Before we get started,
21
00:01:33.079 --> 00:01:34.799
can you tell our listeners give us a bit of
22
00:01:35.239 --> 00:01:38.120
information about your background, about what you know, what you
23
00:01:38.200 --> 00:01:40.400
went up to, and the good work you're doing here
24
00:01:40.400 --> 00:01:41.319
at Omni Security.
25
00:01:41.959 --> 00:01:45.560
Thank you so much, Andrew, and welcome to Norway and
26
00:01:45.680 --> 00:01:48.760
our office. I'm so glad to have you visiting us.
27
00:01:49.560 --> 00:01:53.359
So my name is Kenneth Tiklestad and I'm working as
28
00:01:53.400 --> 00:01:58.799
a chief commercial officer in OMNI and I've just started
29
00:01:58.840 --> 00:02:01.760
as a commercial officer in Omni. I went over from
30
00:02:01.840 --> 00:02:07.519
Supersteria where I was heading up OT cybersecurity. I've been
31
00:02:07.519 --> 00:02:11.240
doing that for six years. Before that, I was working
32
00:02:11.280 --> 00:02:14.759
in ecuinor also working on OT cybersecurity. So I've been
33
00:02:16.039 --> 00:02:19.280
working in the field now for almost fifteen years. And
34
00:02:19.360 --> 00:02:21.759
also for the last five or six years I've been
35
00:02:22.240 --> 00:02:26.159
chairman for the Norwegian Electrotechnical comedy the group that hand
36
00:02:26.840 --> 00:02:28.680
is Handling I C.
37
00:02:29.159 --> 00:02:30.360
Sixty two four four three.
38
00:02:31.159 --> 00:02:34.919
So I've been diving deep into OT cybersecurity now for
39
00:02:35.000 --> 00:02:41.439
quite many years. Yeah, and at Omni we are developing
40
00:02:41.479 --> 00:02:47.639
a software platform for handling cybersecurity and security for critical infrastructure.
41
00:02:48.199 --> 00:02:52.360
It contains us a security knowledge graph and AI that
42
00:02:52.479 --> 00:02:57.319
provides actionable insights into security for critical infrastructure. So it's
43
00:02:57.360 --> 00:03:00.560
about it OT and physical structure.
44
00:03:01.560 --> 00:03:06.800
Our topic today is creditibility. Now this is talking about risk.
45
00:03:08.240 --> 00:03:11.080
You know a lot of people think risk is boring. Okay,
46
00:03:11.120 --> 00:03:13.520
A lot of people when they enter the industrial security space,
47
00:03:13.560 --> 00:03:15.680
they want to know about attacks, they want to know
48
00:03:15.680 --> 00:03:18.439
about the technical bits invites. You tell me that you
49
00:03:18.520 --> 00:03:21.039
got interested in risk a very long time ago. Can
50
00:03:21.080 --> 00:03:23.080
you talk about that? Where where did that come from?
51
00:03:23.439 --> 00:03:28.840
Absolutely, I'm not sure if I when I considered it
52
00:03:28.960 --> 00:03:36.080
as a risk or as as a.
53
00:03:34.520 --> 00:03:35.599
Field of expertise.
54
00:03:36.120 --> 00:03:40.719
So when I was just a small boy, Actually, my
55
00:03:40.840 --> 00:03:44.599
dad he worked as a control room technician offshore at
56
00:03:45.080 --> 00:03:48.280
in Conico Phillips or back then it was called Phillips.
57
00:03:48.919 --> 00:03:52.479
So when I was only two years three years old,
58
00:03:52.520 --> 00:03:56.199
in nineteen seventy seven, he was working at the Bovo
59
00:03:57.560 --> 00:04:00.680
offshore oil and gas platform and I I don't remember
60
00:04:00.719 --> 00:04:04.960
this of course back then, but it was always a
61
00:04:05.000 --> 00:04:08.639
topic around the dinner table at my home where he
62
00:04:08.960 --> 00:04:11.280
talked about how it was working in the oil and
63
00:04:11.280 --> 00:04:15.080
gas business. So in nineteen seventy seven, he was on
64
00:04:15.120 --> 00:04:19.519
his way out to the platform when the big horbo
65
00:04:19.879 --> 00:04:24.120
blowout happened. He was not actually he hadn't arrived at
66
00:04:24.120 --> 00:04:26.519
the platform, but he was on his way out there.
67
00:04:27.000 --> 00:04:30.680
So it really was a big topic around the dinner
68
00:04:30.720 --> 00:04:36.720
table all the time about safety risks involved in oil
69
00:04:36.720 --> 00:04:42.839
and gas. So I was always listening with my small
70
00:04:42.879 --> 00:04:46.279
ears back then, being
Load full transcript
a bit fascinated about this world.
71
00:04:46.360 --> 00:04:49.600
I didn't see the real danger in it, but I
72
00:04:49.720 --> 00:04:53.360
was trying to picture it in my mind what it
73
00:04:53.519 --> 00:04:58.800
was to actually work in these kind of environments. So
74
00:04:59.680 --> 00:05:02.040
I was was kind of primed back when I was
75
00:05:02.319 --> 00:05:06.319
just a small small boy, and later on when I
76
00:05:06.360 --> 00:05:10.920
moved into the I was more into computers, so I
77
00:05:10.959 --> 00:05:14.079
did a lot of gaming and programming on Commodore sixty
78
00:05:14.160 --> 00:05:17.360
four and I started to work in equinor on the
79
00:05:17.399 --> 00:05:21.800
IT side. But I was still fascinated, fascinated about the
80
00:05:21.839 --> 00:05:25.120
core business being oil and gas and production and exploration.
81
00:05:25.720 --> 00:05:29.639
So when I actually got my first trip offshore, I
82
00:05:29.720 --> 00:05:33.000
kind of felt that the circle was closed and I
83
00:05:33.040 --> 00:05:36.439
saw the big world, the industrial world that my dad
84
00:05:36.920 --> 00:05:40.319
had been talking about for several years, and the kind
85
00:05:40.319 --> 00:05:43.480
of the risk perspectives also kicked in the first thing
86
00:05:43.600 --> 00:05:46.040
you meet when your step on board such a platform.
87
00:05:46.160 --> 00:05:49.319
Is the HC focus a lot of focus on HC,
88
00:05:49.879 --> 00:05:53.519
and it's for a reason. And I fully got to
89
00:05:53.600 --> 00:05:57.920
understand that first when I actually came on board such
90
00:05:58.000 --> 00:06:01.680
a facility, I understood why it's so important because it's
91
00:06:01.839 --> 00:06:05.000
it can be really dangerous if you don't have control
92
00:06:05.079 --> 00:06:08.360
over what you're doing. So that's when I actually saw
93
00:06:08.439 --> 00:06:11.920
the big scale of risk as a perspective.
94
00:06:12.639 --> 00:06:15.920
Yeah, offshore platforms are intense. I've never set foot on
95
00:06:15.920 --> 00:06:20.920
one myself, but I've I've heard the stories. It's yeah, yeah, absolutely,
96
00:06:21.079 --> 00:06:26.120
environment and this is I mean we're talking about industrial
97
00:06:26.279 --> 00:06:31.639
cyber security. So you know offshore platforms are intense in
98
00:06:31.720 --> 00:06:35.279
terms of physical risk. Can you talk about cyber risk?
99
00:06:35.720 --> 00:06:41.240
Yeah? Absolutely, it's an emerging topic. So when I was
100
00:06:41.279 --> 00:06:44.680
working in Statle when it was called Statyle now it's Equinor,
101
00:06:45.439 --> 00:06:51.800
we started to look into that area around twenty ten,
102
00:06:51.920 --> 00:06:55.600
twenty and eleven. I still remember the day when people
103
00:06:55.680 --> 00:06:58.639
came charging into the meeting room and they started talking
104
00:06:58.639 --> 00:07:02.360
about the news of stucks net. So that was I
105
00:07:02.360 --> 00:07:05.600
think we got to hear about it. In twenty ten,
106
00:07:05.720 --> 00:07:08.240
I was working on the IT side and I was
107
00:07:08.480 --> 00:07:12.920
responsible for large part parts of our windows infrastructure in
108
00:07:12.959 --> 00:07:16.800
the company, and we started to I started to look
109
00:07:16.839 --> 00:07:21.160
into what this KATA things, what is it. I didn't
110
00:07:21.160 --> 00:07:24.279
know about PLCs. I had never seen a PLC. I
111
00:07:24.319 --> 00:07:27.600
didn't know that there was actually other kind of digital
112
00:07:27.639 --> 00:07:33.879
equipment operating critical infrastructure. So with stucknt, I started to
113
00:07:33.920 --> 00:07:38.240
dive into the landscape of OT cybersecurity and also as
114
00:07:38.240 --> 00:07:42.759
a company, we started a big journey back then on
115
00:07:43.560 --> 00:07:48.920
really making OT much more cybersecurity, and stuck Net was
116
00:07:49.279 --> 00:07:51.160
kind of a kickstart for it.
117
00:07:54.600 --> 00:07:59.360
Andrew, it feels like maybe there are certain kinds of
118
00:07:59.680 --> 00:08:03.120
SEM and all the cybersecurity incidents in the OT world
119
00:08:03.680 --> 00:08:07.000
we talk, we reference often the two thousand and seven
120
00:08:07.040 --> 00:08:12.519
Aurora test, maybe you know Triton and Destroyer, But Stuck's
121
00:08:12.600 --> 00:08:16.920
net is that foundational thing that you know, set the
122
00:08:16.959 --> 00:08:18.600
timeline for everybody.
123
00:08:18.199 --> 00:08:23.959
Right indeed, and you know I was active in the space.
124
00:08:24.000 --> 00:08:27.519
I mean I was leading the team at Industrial Defender
125
00:08:27.879 --> 00:08:31.079
building the world's first industrial SEM at the time, so
126
00:08:31.160 --> 00:08:34.039
stucksnet was big news. I did a lot of work
127
00:08:34.080 --> 00:08:36.279
on stucks Net. I had a blog at the time,
128
00:08:36.399 --> 00:08:38.600
you know, every time I learned something new about it,
129
00:08:38.639 --> 00:08:41.759
because somebody had published a report, somebody had published another blog.
130
00:08:41.799 --> 00:08:44.879
I'd done a little research on my own, I'd published this.
131
00:08:46.240 --> 00:08:51.879
I published a paper on how stucks net spread because
132
00:08:51.919 --> 00:08:55.279
you know, analysis had been done of the artifact, you know,
133
00:08:55.480 --> 00:08:59.120
the malware, but it had been done by it people
134
00:08:59.240 --> 00:09:02.039
at semana tick at, I think E set a bunch
135
00:09:02.039 --> 00:09:05.120
of people had analyzed the malware, and you know, that's
136
00:09:05.159 --> 00:09:07.600
work I couldn't do. I'm not a I'm not a
137
00:09:07.639 --> 00:09:12.679
reverse analyst. But I sat down with Joel langel I,
138
00:09:12.720 --> 00:09:17.919
sat down with Eric Buyers, and we investigated the impact
139
00:09:17.960 --> 00:09:20.480
that stucks net would have in a network. What would
140
00:09:20.559 -->
00:09:22.080
what would happen if you let this thing loose in
141
00:09:22.080 --> 00:09:25.639
a network? Given our understanding of the Semen systems. Joel
142
00:09:25.679 --> 00:09:28.360
was an expert on the Seaman systems, you know, Eric
143
00:09:28.399 --> 00:09:31.279
and I were sort of more expert more generally on
144
00:09:31.279 --> 00:09:36.039
on you know, firewalls and industrial systems. So we all
145
00:09:36.080 --> 00:09:38.799
contributed to this paper and you know, said, here's what
146
00:09:38.799 --> 00:09:42.480
happens if you let loose stucks net into an industrial network.
147
00:09:44.080 --> 00:09:47.679
And you know, in hindsight, I have to wonder if
148
00:09:47.879 --> 00:09:51.039
we didn't do you know, more damage than than good.
149
00:09:51.600 --> 00:09:54.120
You know, because a lot of people learned stuff about
150
00:09:54.120 --> 00:09:57.720
sucks net, but there was only one outfit that benefited,
151
00:09:57.840 --> 00:10:00.799
and that was Iran's nuclear weapons program was the only
152
00:10:01.440 --> 00:10:03.840
you know, site in the world that was physically impacted.
153
00:10:03.879 --> 00:10:06.840
So I, you know, I regret some of the stuff
154
00:10:06.879 --> 00:10:08.679
that I published about about stocksnet.
155
00:10:09.559 --> 00:10:12.799
Do you recall if that research got traction, whether it
156
00:10:12.919 --> 00:10:14.679
might have gotten over there or is there no way
157
00:10:14.720 --> 00:10:15.080
to tell?
158
00:10:15.480 --> 00:10:19.440
I have no way to tell. I do recall a conversation,
159
00:10:19.600 --> 00:10:22.799
you know, sometime later, you know, because I'm a Canadian,
160
00:10:22.919 --> 00:10:26.799
I work with the Canadian authorities. I remember a conversation
161
00:10:26.879 --> 00:10:32.840
with Canadian Intelligence services and I remember, you know, asking them,
162
00:10:33.279 --> 00:10:37.519
you know, I've I've stopped, you know, at one point,
163
00:10:38.039 --> 00:10:40.759
when I figured out that there's only one place in
164
00:10:40.799 --> 00:10:44.720
the world that's physically benefiting from my research, I stopped
165
00:10:44.759 --> 00:10:49.240
publishing anything about stucksnet. And I remember sometime after that
166
00:10:49.399 --> 00:10:52.879
talking to Canadian intelligence saying, you know, I've stopped publishing
167
00:10:52.879 --> 00:10:55.000
anything about stocks neet. You don't have to tell me
168
00:10:55.039 --> 00:10:59.480
nothing in the future. If you ever see me putting
169
00:10:59.559 --> 00:11:04.039
out in that's helping our enemies, tap me on the shoulder,
170
00:11:04.039 --> 00:11:06.720
would you and tell me shut up? Ginter, you're doing
171
00:11:06.759 --> 00:11:09.960
more harm than good and I will shut up. So yeah,
172
00:11:10.000 --> 00:11:12.279
I you know, I look back on stucks net with
173
00:11:12.679 --> 00:11:15.200
mixed emotions. It was a wake up call for the industry.
174
00:11:15.519 --> 00:11:17.639
You know a lot of people learned about cybersecurity because
175
00:11:17.639 --> 00:11:23.960
the stucks net. But who benefited because of all that research? Okay,
176
00:11:24.000 --> 00:11:25.919
so that's you know, stucks net is how a lot
177
00:11:25.919 --> 00:11:27.840
of people got started in the AT space. It was
178
00:11:27.919 --> 00:11:33.799
the big news. Yeah, fifteen fifteen years ago. Can I
179
00:11:33.840 --> 00:11:36.840
ask you, you know, let's let's talk about industrial security
180
00:11:36.879 --> 00:11:40.879
and the work you're doing. You the work you've been doing.
181
00:11:42.240 --> 00:11:44.600
Stucks net is where it got started. Where have you
182
00:11:44.600 --> 00:11:46.080
wound up? What are you up to today?
183
00:11:46.919 --> 00:11:50.279
Yeah, it's it's as you say, it's fifteen years and
184
00:11:50.360 --> 00:11:53.759
it's been for me. I think it's been a very
185
00:11:53.759 --> 00:11:58.159
interesting journey. So but back in twenty ten, when when
186
00:11:58.240 --> 00:12:02.600
stucks net hit the news, I wasn't immediately immediately diving
187
00:12:02.600 --> 00:12:06.200
into OT cybersecurity full time. I was working on the
188
00:12:06.240 --> 00:12:11.080
IT side, trying to secure windows environment in a large
189
00:12:11.080 --> 00:12:16.159
oil and gas company. But shortly after a while I
190
00:12:16.320 --> 00:12:19.320
moved more and more over to OT cybersecurity, and I
191
00:12:19.360 --> 00:12:22.919
had my first trip offshore oil and gas platform. I
192
00:12:22.960 --> 00:12:25.799
think that first trip was in twenty thirteen, so actually
193
00:12:25.879 --> 00:12:28.399
three years after this dug Net. But then I was
194
00:12:28.440 --> 00:12:31.679
going out just to do some troubleshooting on a firewall.
195
00:12:32.559 --> 00:12:37.639
So but more and more I was moving into OT cybersecurity,
196
00:12:37.960 --> 00:12:40.559
and at the end I was I moved over to
197
00:12:40.600 --> 00:12:44.279
Supersteria I think it was in twenty seventeen, and at
198
00:12:44.279 --> 00:12:49.799
the end I was really working hard on finding really
199
00:12:50.000 --> 00:12:56.799
proper solutions for OT cybersecurity. When when potential nation states
200
00:12:56.840 --> 00:13:01.120
are targeting, what do you then do if you must
201
00:13:01.440 --> 00:13:05.600
sort of have their mindset of assume breach and these
202
00:13:05.679 --> 00:13:08.240
kind of systems with the PLCs and or they are
203
00:13:08.279 --> 00:13:10.879
really really vulnerable, what do you do when you are
204
00:13:11.120 --> 00:13:17.200
being targeted? So then I started to look into I
205
00:13:17.240 --> 00:13:20.240
heard rumors that could there could be something that was
206
00:13:20.320 --> 00:13:25.360
non hackable, so I started investigating into UNI directional data
207
00:13:25.440 --> 00:13:29.399
diotes was exposed to Waterfall. That was one of the
208
00:13:29.440 --> 00:13:33.240
first
examples of where I heard about non hackable stuff.
209
00:13:34.039 --> 00:13:39.799
And also I got to hear about the Crown Jewel
210
00:13:39.919 --> 00:13:45.240
analysis cyber informed engineering. Back then it was consequence driven
211
00:13:45.360 --> 00:13:49.600
cyber informed engineering. But those kind of topics really really
212
00:13:50.080 --> 00:13:52.919
sparked an extra interest for me because then I saw
213
00:13:54.159 --> 00:13:57.240
on some attack vectors, on some of the risks. I
214
00:13:57.279 --> 00:14:00.919
saw actually a solution that could remove risk instead of
215
00:14:01.039 --> 00:14:02.120
just mitigating it.
216
00:14:02.840 --> 00:14:06.720
So your first sort of for a you know, everyone
217
00:14:07.080 --> 00:14:09.639
was interested in stock step, but you started working on
218
00:14:09.679 --> 00:14:13.679
the problem. You said, with a firewall, And you know,
219
00:14:13.960 --> 00:14:16.879
to a degree that makes sense. I mean the firewall,
220
00:14:17.000 --> 00:14:21.399
the ITOT firewalls often the boundary between the engineering discipline
221
00:14:22.000 --> 00:14:26.240
on the platform in the industrial process and the IT discipline,
222
00:14:26.279 --> 00:14:29.240
where information is the asset that needs to be protected,
223
00:14:30.519 --> 00:14:33.759
and so that boundary is something that both the engineers
224
00:14:33.799 --> 00:14:37.519
and the IT folk care about. So that kind of
225
00:14:37.519 --> 00:14:40.360
makes sense. I'm curious. You know, you got out to
226
00:14:40.440 --> 00:14:43.679
the platform, you were tasked with the firewall. What did
227
00:14:43.720 --> 00:14:44.360
you find there?
228
00:14:44.840 --> 00:14:49.480
Yeah, it was actually kind of a long long lasting
229
00:14:49.519 --> 00:14:51.960
ticket we had in our system that was a firewall
230
00:14:52.039 --> 00:14:55.279
between IT and OT that was noisy. So it was
231
00:14:56.120 --> 00:15:01.679
creating a lot of events and alert on traffic that
232
00:15:01.879 --> 00:15:05.559
it shouldn't have. So I was tasked to go out
233
00:15:05.559 --> 00:15:10.320
there and try to trouble shoot this. We we absolutely
234
00:15:10.399 --> 00:15:13.679
didn't think that it was a cyber attack or or
235
00:15:13.759 --> 00:15:18.279
kind of evil intent, but it was incorrectly configured firewall rule.
236
00:15:18.639 --> 00:15:22.519
But when I got out there, I could see that
237
00:15:22.720 --> 00:15:30.480
it was it was just incorrectly configured fire world was nothing,
238
00:15:30.159 --> 00:15:35.440
not anything dangerous or a cyber attack involved. But I
239
00:15:35.559 --> 00:15:40.240
also got to think of a scenario where if it
240
00:15:40.360 --> 00:15:44.879
had actually been a cyber attack and one that created
241
00:15:45.000 --> 00:15:48.399
so much noise as well on a security boundary, a
242
00:15:48.480 --> 00:15:54.360
security component sitting on the outskirts of OT, shouldn't the
243
00:15:54.399 --> 00:15:57.919
OT environment do something to sort of shut down or
244
00:15:58.000 --> 00:16:00.399
go into a more failed safe situation. So I got
245
00:16:00.639 --> 00:16:05.679
kind of interesting in actually the instrumentation behind your security
246
00:16:05.759 --> 00:16:11.120
components on the outskirts of OT. So that's a topic
247
00:16:11.240 --> 00:16:15.559
I continued to explore for several years, having in the
248
00:16:15.600 --> 00:16:20.240
back of my mind cyber informed engineering, non hackabal approaches,
249
00:16:20.399 --> 00:16:26.559
unidirectional mechanisms, And on S four last year, I talked
250
00:16:26.559 --> 00:16:31.720
about the safety instrumented system because safety has always been
251
00:16:31.720 --> 00:16:37.159
a particular interest of mine, so I talked about the
252
00:16:37.320 --> 00:16:43.159
cyber informed safety instrumental system. Shouldn't the safety instrumented system
253
00:16:43.600 --> 00:16:47.279
at some point when you're under an attack, shouldn't the
254
00:16:48.360 --> 00:16:51.039
sort of the big brain in the room, shouldn't that
255
00:16:51.120 --> 00:16:55.799
actually take an action, an instrumented automated action, and going
256
00:16:55.840 --> 00:16:59.879
into a more not necessarily failed safe only, but a
257
00:17:00.200 --> 00:17:04.799
more fail fail over to a more safe and secure situation.
258
00:17:05.839 --> 00:17:08.319
So that makes sense in theory. I mean, if the
259
00:17:08.359 --> 00:17:11.319
firewall was saying help, help, I'm under attack over and
260
00:17:11.359 --> 00:17:15.240
over again, should some action not have taken place on
261
00:17:15.279 --> 00:17:17.920
the OT side. But let me ask you this, it
262
00:17:18.039 --> 00:17:20.640
was a false positive. Yes, it would have shut down
263
00:17:20.640 --> 00:17:26.079
the platform, you know, a very expensive platform unnecessarily. Can
264
00:17:26.160 --> 00:17:32.279
we detect cyber attacks reliably enough to prevent this kind
265
00:17:32.319 --> 00:17:37.440
of unnecessary shutdown? You know? And you know if we
266
00:17:37.519 --> 00:17:40.519
do shut down whenever there's a bunch of alarms, is
267
00:17:40.519 --> 00:17:44.119
that not a new sort of denial of service vulnerability?
268
00:17:44.119 --> 00:17:46.400
The bad guys don't even need to get into OT.
269
00:17:46.960 --> 00:17:49.359
They just need to launch a few package that firewall,
270
00:17:49.440 --> 00:17:52.279
generate some alarms, and the whole thing shuts down without
271
00:17:52.319 --> 00:17:55.559
them even bothering to break into OT. Is that really
272
00:17:55.599 --> 00:17:57.240
the right way forward? No?
273
00:17:57.720 --> 00:18:01.279
I totally agree, it's not a good coach going forward.
274
00:18:02.720 --> 00:18:05.640
But at the same time, I think to shut down
275
00:18:05.799 --> 00:18:09.680
one too many times is better than not actually doing it.
276
00:18:09.960 --> 00:18:16.880
So we
should be kind of overreacting and going into
277
00:18:16.920 --> 00:18:22.039
failed safe situation and it could cause unnecessary downtime and
278
00:18:22.079 --> 00:18:25.839
it could it's the vulnerability on the production side, but
279
00:18:25.920 --> 00:18:29.599
I think it's much more dangerous with the false negatives
280
00:18:30.160 --> 00:18:35.119
where we actually don't see any attacks, but it's it's
281
00:18:35.359 --> 00:18:38.920
actually happening, so false positive. We need to reduce them.
282
00:18:38.960 --> 00:18:42.480
But it's much more important to actually reduce the false negatives.
283
00:18:45.720 --> 00:18:48.319
So Nate just listening to the recording here. I mean,
284
00:18:48.319 --> 00:18:50.640
this is not something I discussed with Kenneth, but we
285
00:18:50.640 --> 00:18:55.799
were talking about, you know, automatic action when we discovered
286
00:18:55.799 --> 00:18:57.960
that an attack might be in progress, for example, because
287
00:18:57.960 --> 00:18:59.960
there's a lot of alarms coming out of the firewall.
288
00:19:01.640 --> 00:19:05.519
You know, he agreed with me that shutting down the
289
00:19:05.559 --> 00:19:09.680
platform is probably an overreaction because you know, that introduces
290
00:19:09.720 --> 00:19:11.759
a new attack vector. The bad guys just need to
291
00:19:11.799 --> 00:19:14.680
send a few packets against the firewall, generate a few alarms,
292
00:19:14.839 --> 00:19:18.319
and the whole platform shuts down. I agreed with him
293
00:19:18.359 --> 00:19:22.319
that something should be done, but we didn't really figure
294
00:19:22.319 --> 00:19:26.880
out what. You know, here's an idea. In hindsight, A
295
00:19:26.960 --> 00:19:31.240
number of jurisdictions are introducing what they call islanding rules,
296
00:19:31.559 --> 00:19:37.119
meaning if it is compromised, you need to you know, basically,
297
00:19:37.200 --> 00:19:40.519
I don't know, power off the it OT firewall, nothing
298
00:19:40.880 --> 00:19:45.039
gets through into OT anymore for the duration of the emergency,
299
00:19:45.319 --> 00:19:49.680
so you have the ability to shut off all communications
300
00:19:49.720 --> 00:19:52.519
into OT. This is part of you know, the regulation
301
00:19:52.599 --> 00:19:54.359
says you must be able to island, So now you
302
00:19:54.400 --> 00:19:57.240
have that capability. You know, I wonder if it isn't
303
00:19:57.359 --> 00:20:04.359
reasonable to trigger islanding when when you discover, you know,
304
00:20:04.480 --> 00:20:07.200
automatically discover a whole bunch of alarms coming out of anything.
305
00:20:07.240 --> 00:20:10.559
Because the modern attack pattern most of them of modern
306
00:20:10.640 --> 00:20:12.440
day attacks are not like stocks net where you let
307
00:20:12.480 --> 00:20:14.160
it loosen, it does its things. Most of modern day
308
00:20:14.200 --> 00:20:17.039
attacks have remote control from the Internet. And if you island,
309
00:20:17.039 --> 00:20:20.519
if you break the connection between it and OT, if
310
00:20:20.599 --> 00:20:23.319
there was an attack in the OT network, the bad
311
00:20:23.359 --> 00:20:25.759
guys can no longer control it, they can no longer
312
00:20:25.799 --> 00:20:29.160
send commands. So and this is not this is not new.
313
00:20:29.200 --> 00:20:32.240
The term islanding is a little bit new. The concept
314
00:20:32.359 --> 00:20:35.839
of sort of an automatic shutoff has been bandied about
315
00:20:35.839 --> 00:20:39.880
for many years. But again, given that the regulators are
316
00:20:39.880 --> 00:20:44.680
demanding an islanding capability, you know, maybe engaging it automatically
317
00:20:44.680 --> 00:20:47.640
from time to time is not the worst thing that
318
00:20:47.640 --> 00:20:51.000
can happen. It increases our security and the impact on
319
00:20:51.079 --> 00:20:56.920
operations is minimal because you've you've deployed the ability to
320
00:20:57.079 --> 00:21:02.279
island already, you've developed the capability of running EUROT system
321
00:21:02.319 --> 00:21:06.400
independently and so uh, you know, interrupting that communication for
322
00:21:06.440 --> 00:21:08.359
a period of hours at a time while you track
323
00:21:08.400 --> 00:21:11.160
things down and say, oh, that was a false alarm.
324
00:21:12.079 --> 00:21:17.079
I'm guessing is you know, minimal cost. So there's an idea.
325
00:21:18.880 --> 00:21:22.440
Let's come back to our topic here. The topic is credibility.
326
00:21:24.000 --> 00:21:27.720
You know, we're talking about the risk equation. The typical
327
00:21:27.799 --> 00:21:33.559
risk equation is consequence times likelihood. Uh, you know, generally
328
00:21:33.559 --> 00:21:36.200
we do it qualitatively, but we we wind up with
329
00:21:36.240 --> 00:21:39.799
a number coming out of that to compare different different
330
00:21:39.880 --> 00:21:43.400
kinds of risks, you know, high frequency versus versus high
331
00:21:43.440 --> 00:21:49.720
impact risks. You know, can you talk about that? Where
332
00:21:49.720 --> 00:21:52.799
does credibility fit in that equation?
333
00:21:53.519 --> 00:21:57.279
I think it fits very well into that equation because
334
00:21:59.039 --> 00:22:01.920
when we when we talk about it likelihood or the
335
00:22:01.920 --> 00:22:04.839
probability part of it, the left side of the equation.
336
00:22:05.319 --> 00:22:11.000
It's always a very very difficult conversation to have when
337
00:22:11.039 --> 00:22:15.480
you try to identify the risk or the risk levels
338
00:22:15.480 --> 00:22:18.240
we are talking about, or you try to identify the
339
00:22:18.319 --> 00:22:23.279
consequence levels involved. It's sad to see that a lot
340
00:22:23.279 --> 00:22:27.240
of the conversations they go astray due to not being
341
00:22:27.319 --> 00:22:30.759
able to put the number on the probability or the likelihood.
342
00:22:31.240 --> 00:22:35.319
And
I think the conversation gets to be much more
343
00:22:35.319 --> 00:22:39.079
fruitful if we can get rid of that challenge on
344
00:22:39.160 --> 00:22:41.640
trying to figure out the number on the probability or
345
00:22:41.640 --> 00:22:48.039
the likelihood. Credibility gives us tools in our language to
346
00:22:48.160 --> 00:22:51.559
actually be able to talk about the left part of
347
00:22:51.640 --> 00:22:56.200
the equation. So it's something that is a bit more
348
00:22:56.240 --> 00:23:02.000
analog and analog value where we can moved more towards
349
00:23:02.039 --> 00:23:06.839
the consequence approach, the consequence driven where the right side
350
00:23:06.839 --> 00:23:10.519
of the equation is more important to talk about. As
351
00:23:10.599 --> 00:23:13.960
long as you get if you consider it being credible,
352
00:23:14.359 --> 00:23:19.319
then okay, let's stop the discussion there and focus on
353
00:23:19.440 --> 00:23:21.440
identify the consequence levels.
354
00:23:22.160 --> 00:23:25.960
Well, you know I have to agree. You know, I've
355
00:23:26.319 --> 00:23:31.480
argued in my previous in my last book, that likelihood
356
00:23:31.519 --> 00:23:35.960
is flawed, that the high end of cyber attacks, not
357
00:23:36.000 --> 00:23:38.799
the low end. The low end. Likelihood actually works the
358
00:23:38.920 --> 00:23:43.599
high end. The outcomes of cyber attacks are not random.
359
00:23:43.880 --> 00:23:47.160
If the same ransomware hits a factory twice, and we've
360
00:23:47.160 --> 00:23:49.400
always done is restore from backrupt. It took them down
361
00:23:49.480 --> 00:23:52.839
the first time. We restore from backup, we make no changes.
362
00:23:53.559 --> 00:23:55.839
It hits again, They're going to go down the same way.
363
00:23:55.920 --> 00:23:59.960
It's not random. I argue that on the high end,
364
00:24:00.519 --> 00:24:04.759
Nation state targeting is not random either. You know, it's
365
00:24:04.759 --> 00:24:07.720
not that they try for a while and they if
366
00:24:07.759 --> 00:24:10.200
they don't succeed, they you know, go try somewhere else.
367
00:24:11.799 --> 00:24:17.519
Nation state threat actors keep targeting the same target until
368
00:24:17.519 --> 00:24:20.880
they achieve their mission objective. It's not random. Once they've
369
00:24:20.920 --> 00:24:25.720
targeted you, it's not random. So you know, randomness, to me,
370
00:24:25.799 --> 00:24:32.119
doesn't work at the high end. Credibility makes more sense.
371
00:24:32.200 --> 00:24:35.440
You know, is the threat credible? Is the consequence credible?
372
00:24:35.440 --> 00:24:37.559
If this threat comes after us, it's this attack comes
373
00:24:37.559 --> 00:24:40.920
after us. Is it reasonable to believe? You know, credibility
374
00:24:41.000 --> 00:24:44.119
is what's reasonable to believe, not who what's reasonable to believe?
375
00:24:44.160 --> 00:24:46.720
Is it reasonable to believe that the consequence will be realized?
376
00:24:48.799 --> 00:24:50.839
You know, I think it makes a lot of sense,
377
00:24:50.880 --> 00:24:54.440
but it's it's new. I don't see the word credibility
378
00:24:54.440 --> 00:25:01.000
in a lot of standards. You know, where does this it? What? What
379
00:25:01.559 --> 00:25:05.319
you know? Is this is this something people are talking about.
380
00:25:06.480 --> 00:25:06.720
Yeah.
381
00:25:06.759 --> 00:25:07.400
Absolutely.
382
00:25:08.920 --> 00:25:12.799
In my work with with the clients I've been working
383
00:25:12.839 --> 00:25:15.759
with and also the professionals I've been working with, we
384
00:25:15.880 --> 00:25:19.839
have discussed for some years now that the or we
385
00:25:19.920 --> 00:25:23.039
have discussed the big challenge of the likelihood or the
386
00:25:23.039 --> 00:25:29.000
probability part of the equation, and we've we've without actually
387
00:25:29.039 --> 00:25:35.759
having having without following standards or best practices, we've seen
388
00:25:35.839 --> 00:25:39.640
that we need to skip the discussion on the probability
389
00:25:39.759 --> 00:25:42.559
or the likelihood and talk about the consequence side of
390
00:25:42.599 --> 00:25:47.079
it first, and then we revisit the likelihood and probability afterwards.
391
00:25:47.319 --> 00:25:50.000
But I also see in ie C sixty two four three,
392
00:25:51.200 --> 00:25:56.119
especially with three dash two, it actually talks about consequence
393
00:25:56.240 --> 00:26:05.960
only risk analysis, So that's giving opportunity to actually move
394
00:26:06.000 --> 00:26:09.759
away from the discussions on probability. And also, of course
395
00:26:09.799 --> 00:26:13.640
with the consequence driven approach with cyber informed engineering, we
396
00:26:13.720 --> 00:26:17.039
start to see more focus on the far right side
397
00:26:17.079 --> 00:26:21.519
with the consequence consequence side, but leaving out.
398
00:26:22.759 --> 00:26:24.599
What to do with the likelihood.
399
00:26:24.880 --> 00:26:29.519
And I think with credibility we get some language based
400
00:26:29.599 --> 00:26:34.519
tools to actually place it where we talk about it
401
00:26:34.720 --> 00:26:38.200
in a qualitative manner instead of having to force it
402
00:26:38.599 --> 00:26:39.599
into a number.
403
00:26:40.799 --> 00:26:45.519
I have the sense that over time, in the course
404
00:26:45.559 --> 00:26:51.519
of time, cyber attacks become more sophisticated, More sophisticated attacks
405
00:26:51.559 --> 00:26:54.759
become credible. Attacks that were dismissed a decade ago as
406
00:26:54.960 --> 00:26:59.279
theoretical have actually happened. Do you see that, you know,
407
00:26:59.319 --> 00:27:02.279
what do you see coming at us in terms of
408
00:27:02.519 --> 00:27:04.960
sophisticated attacks in the near future?
409
00:27:05.000 --> 00:27:10.640
Here, I think that's a really challenging question looking far
410
00:27:10.720 -->
00:27:14.480
into the future or or far into the into the
411
00:27:14.559 --> 00:27:19.519
history to try to extrapolate what could we expect from
412
00:27:19.559 --> 00:27:20.079
the future.
413
00:27:20.480 --> 00:27:21.640
We see with with.
414
00:27:21.759 --> 00:27:29.599
The stocksnet, the attacks against Ukraine, Triton, the colonial pipeline,
415
00:27:30.039 --> 00:27:34.720
we see incidents that have had a really high impact,
416
00:27:35.160 --> 00:27:38.960
but there's not very many of them. So but but
417
00:27:39.119 --> 00:27:46.039
we see it's those kind of capabilities are being explored
418
00:27:46.359 --> 00:27:50.759
and are being put into different tools so they can
419
00:27:50.799 --> 00:27:55.559
be used by not only nation states but also criminal groups.
420
00:27:56.279 --> 00:28:00.359
So with with that kind of analysis, we can expect
421
00:28:01.039 --> 00:28:05.720
more and more sophisticated attacks and also buy more and
422
00:28:05.880 --> 00:28:10.720
more non sophisticated groups. So we should expect increase in
423
00:28:11.799 --> 00:28:13.519
high impact incidents.
424
00:28:13.519 --> 00:28:16.039
So if we're not talking likelihood, we're not talking probability,
425
00:28:16.279 --> 00:28:20.400
we're talking credible. How do we decide what's credible? How
426
00:28:20.400 --> 00:28:22.519
do we decide what's reasonable to believe?
427
00:28:24.000 --> 00:28:29.440
Yeah, that's a good question. So we need to have
428
00:28:29.559 --> 00:28:34.039
some grasp of what is credible and what is not credible.
429
00:28:34.640 --> 00:28:42.160
I'm also of the opinion that the credibility part of
430
00:28:42.200 --> 00:28:47.200
the equation. It's a qualitative thing. It's not a zero
431
00:28:47.440 --> 00:28:51.640
or one. It's something that is attached to a kind
432
00:28:51.680 --> 00:28:56.400
of a slippery slope, not easily defined. But what we
433
00:28:56.440 --> 00:29:02.599
could say, if we are trying to see credibility as
434
00:29:02.599 --> 00:29:06.640
a zero or one, what is credible? Things that have happened,
435
00:29:06.880 --> 00:29:10.319
actually have happened once or twice or three times, they
436
00:29:10.359 --> 00:29:16.400
are credible. So the Triton incident or a safety only
437
00:29:16.759 --> 00:29:21.880
type of cybersecurity attack, that's now a credible attack because
438
00:29:21.920 --> 00:29:25.640
it has happened. And also near misses. That's something that
439
00:29:26.200 --> 00:29:28.759
Triton was kind of a near miss. They didn't actually
440
00:29:28.880 --> 00:29:35.000
cause this destructive attack, but it could have happened, and
441
00:29:35.079 --> 00:29:40.680
so we also have other near missus incidents that we
442
00:29:41.079 --> 00:29:44.799
should be considering as credible attacks.
443
00:29:45.519 --> 00:29:49.119
Credibility sounds like a judgment called how do we decide what's.
444
00:29:48.960 --> 00:29:55.920
Credible's that's a good question. I think there's a good
445
00:29:56.960 --> 00:30:00.680
recommendations in sixty two four for three for instance, three
446
00:30:00.759 --> 00:30:06.960
thatsh two. It talks about, like I said, the consequence
447
00:30:07.079 --> 00:30:09.839
only as an example on how how you can approach
448
00:30:09.920 --> 00:30:14.480
the risk equation. But it also talks about the need
449
00:30:14.680 --> 00:30:20.200
for focusing on worse case consequences. So there's it talks
450
00:30:20.240 --> 00:30:23.920
about essential functions, which basically it could be the safety
451
00:30:23.920 --> 00:30:29.079
functions for instance, you need to investigate the consequence if
452
00:30:29.119 --> 00:30:33.200
those are are actually attacked and compromised, what could be
453
00:30:33.240 --> 00:30:37.000
the worst case consequence? So you begin there and then
454
00:30:37.359 --> 00:30:41.039
once you identify the worst case consequences, then you move
455
00:30:41.079 --> 00:30:47.440
over to the probability or likelihood dimension. And then then
456
00:30:47.480 --> 00:30:49.000
you need to consider other factors.
457
00:30:49.480 --> 00:30:50.319
So what are the.
458
00:30:50.359 --> 00:30:54.960
Vulnerabilities involved, what are the safeguards and or what the
459
00:30:55.000 --> 00:30:59.400
standard is talking about your compensating countermeasures, you consider that
460
00:31:00.119 --> 00:31:04.640
you consider the function or the asset as well that
461
00:31:05.319 --> 00:31:09.279
if there's if there's no actual interest in the asset,
462
00:31:09.440 --> 00:31:16.119
then the vulnerability could be also non interesting to address
463
00:31:16.680 --> 00:31:21.839
or analyz. So but you start with the consequence side.
464
00:31:22.319 --> 00:31:24.759
Then you start to look at the likelood and probability,
465
00:31:26.359 --> 00:31:31.200
and you are informed by the consequence approach.
466
00:31:31.920 --> 00:31:34.799
Okay, so let me challenge you on that. I've read
467
00:31:34.960 --> 00:31:39.039
the c IE Implementation Guide. It says start with the
468
00:31:39.079 --> 00:31:43.279
worst case consequences. It says those words. I've not seen
469
00:31:43.319 --> 00:31:46.079
those words in three dash two. Are you sure that
470
00:31:46.400 --> 00:31:49.160
you're not reading in the three dash too to be there.
471
00:31:49.640 --> 00:31:52.720
I've been searching for that specific part of three dish
472
00:31:52.759 --> 00:31:56.200
too many times because because I've heard others say that
473
00:31:56.359 --> 00:32:00.799
the same, and it's actually there. It's really gold nuggets
474
00:32:00.839 --> 00:32:05.200
in three dish two talking about the essential functions, specifically
475
00:32:05.240 --> 00:32:09.400
saying the worst case consequence and also specifically saying that
476
00:32:09.519 --> 00:32:13.400
you can choose to do a consequence only risk assessment.
477
00:32:13.559 --> 00:32:20.160
So that's really really important.
Single words or single sentences
478
00:32:20.279 --> 00:32:24.880
in three dish two so worth highlighting in the three
479
00:32:24.960 --> 00:32:25.359
dish two.
480
00:32:26.440 --> 00:32:29.480
Okay, so that makes sense in the abstract. Can you
481
00:32:29.519 --> 00:32:33.519
give me some examples what you know applying these principles,
482
00:32:33.680 --> 00:32:36.240
what should we regard as credible?
483
00:32:36.960 --> 00:32:41.039
Yeah? Interesting question. I think that the things that come
484
00:32:41.079 --> 00:32:48.720
to mind first is, for instance, the Triton incident before
485
00:32:48.839 --> 00:32:51.880
twenty seventeen, where when it actually happened, we didn't think
486
00:32:52.000 --> 00:32:55.759
it was credible that someone would actually target a safety
487
00:32:55.799 --> 00:32:59.240
only system or cause a safety incident with a cyber
488
00:32:59.240 --> 00:33:04.240
attack with Triton it we actually saw the first first
489
00:33:04.279 --> 00:33:10.720
of its kind and the threat became obviously credible. And
490
00:33:10.759 --> 00:33:14.519
then Solo Winds as well. It's a very interesting study
491
00:33:15.000 --> 00:33:19.160
where the way they actually compromised the Solo Winds update
492
00:33:19.319 --> 00:33:30.240
mechanism suddenly massive, massive deployment of kind of malware within
493
00:33:30.759 --> 00:33:35.440
critical and non critical infrastructure became a really credible threat
494
00:33:35.599 --> 00:33:36.039
as well.
495
00:33:36.559 --> 00:33:37.680
And also near misses.
496
00:33:37.720 --> 00:33:41.079
Of course, we should be informed by things happening out
497
00:33:41.119 --> 00:33:45.480
there and coming on the news that are near misses
498
00:33:45.640 --> 00:33:48.200
that can talk about talk to us about what is
499
00:33:48.240 --> 00:33:51.799
a credible threat. Another kind of near miss that I
500
00:33:51.839 --> 00:33:56.480
think or it's not a near miss, but it's scenarios
501
00:33:56.599 --> 00:34:00.519
or incidents that could talk about credibility is is where
502
00:34:00.559 --> 00:34:05.480
we actually have a safety incident. For instance, we have
503
00:34:06.200 --> 00:34:08.519
had lots of them in Norwegian oil and gas and
504
00:34:08.559 --> 00:34:12.719
in oil and Jazz gas in general. Is safety incidents
505
00:34:12.800 --> 00:34:16.639
where we which is not cyber related at all, but
506
00:34:16.760 --> 00:34:20.199
where we see that it's it could be able to
507
00:34:20.280 --> 00:34:24.039
be replicated by a cyber attack. So that's something that
508
00:34:24.079 --> 00:34:28.519
we should be considering as a credible threat going forward,
509
00:34:28.840 --> 00:34:33.840
where we actually could replicate the cyber or the incident
510
00:34:34.320 --> 00:34:38.960
with the cyber course on credibility, I also think we
511
00:34:39.000 --> 00:34:43.440
need to put have in the back of our mind
512
00:34:43.559 --> 00:34:46.840
or in the analysis, we have to have focus on
513
00:34:47.280 --> 00:34:51.719
the technology evolution to development and sharing of new technology.
514
00:34:52.039 --> 00:34:55.599
So I see it as a graph where where we
515
00:34:55.639 --> 00:35:00.760
are exposed to more and more heavy machine or heavy
516
00:35:01.800 --> 00:35:05.480
software that can be used on the adversary side, So
517
00:35:05.639 --> 00:35:10.639
with Kali, Li Nux, metasploit nowadays also AI. So what
518
00:35:10.760 --> 00:35:15.480
is being becoming a credible threat threat is more and
519
00:35:15.559 --> 00:35:22.159
more sophisticated stuff due to development of technology. So AI
520
00:35:22.480 --> 00:35:26.599
now is on both sides of the table, both as
521
00:35:26.639 --> 00:35:31.840
an attacker, as a tool that makes more more attacks credible,
522
00:35:32.480 --> 00:35:34.599
but also on the on the defensive side where we
523
00:35:34.679 --> 00:35:40.679
actually need to use it to to protect against more
524
00:35:40.679 --> 00:35:42.480
and more sophisticated attacks.
525
00:35:45.320 --> 00:35:47.199
Let me go just a little bit deeper into into
526
00:35:47.280 --> 00:35:50.079
Kenneth's last example. I remember talking to him about this
527
00:35:51.039 --> 00:35:57.119
two days before I recorded the session with Kenneth. I
528
00:35:57.280 --> 00:36:01.960
was at another event, you know, I had a half
529
00:36:02.039 --> 00:36:04.800
hour speaking slot. I was, you know, listening politely to
530
00:36:04.840 --> 00:36:08.000
the other speakers, I remember, and one of the speakers
531
00:36:08.039 --> 00:36:10.599
was a penetration tester. I remember asking the pen tester
532
00:36:10.719 --> 00:36:14.280
a question about AI and his answer alarmed me. And
533
00:36:14.320 --> 00:36:16.679
you know, I discussed it with Kenneth. I disgusted with
534
00:36:16.840 --> 00:36:22.039
other people, since you know, the future is is difficult.
535
00:36:23.440 --> 00:36:26.159
I asked the AI you know, the pen tester, so
536
00:36:26.239 --> 00:36:28.639
you know you touched on AI. What should we look
537
00:36:28.679 --> 00:36:31.519
for from AI going forward? And I asked, you know,
538
00:36:31.559 --> 00:36:35.519
should we worry about about AI crafting phishing attacks because
539
00:36:35.800 --> 00:36:37.679
I've heard of that happening. Should we worry about AI
540
00:36:37.880 --> 00:36:41.440
helping the bad guys write malware to write more sophisticated malware,
541
00:36:41.440 --> 00:36:44.199
because I've heard of that happening, you know. And I paused,
542
00:36:44.679 --> 00:36:47.639
and his answer was, Andrew, you're not thinking hard enough
543
00:36:47.639 --> 00:36:51.000
about this problem. You know. Yeah, that stuff's happening. But
544
00:36:51.079 --> 00:36:55.079
what you need to worry about is somebody taking a
545
00:36:55.159 --> 00:36:59.400
Cali limit Linux ISO image. This is the Linux disc
546
00:36:59.440 --> 00:37:02.719
image that every body uses. All
the pen testers use
547
00:37:03.400 --> 00:37:10.000
lots of attack tools, he says, taking that gigabyte of isoimage,
548
00:37:10.440 --> 00:37:13.719
you know, coupling it, adding it together with two gigabytes
549
00:37:13.880 --> 00:37:17.480
of AI. Model and the model has not been trained
550
00:37:17.480 --> 00:37:20.599
on natural language and creating phishing attacks. The model has
551
00:37:20.639 --> 00:37:27.400
been trained by watching professional pen testers attack OT systems
552
00:37:27.800 --> 00:37:29.480
mostly in test beds. I mean, this is what pen
553
00:37:29.559 --> 00:37:32.360
testers do. They take a test bed that is a
554
00:37:32.400 --> 00:37:34.519
copy of a system that they're supposed to be, you know,
555
00:37:34.840 --> 00:37:36.679
doing the pen test on No one that does the
556
00:37:36.679 --> 00:37:38.280
pen test on a live system. They do it on
557
00:37:38.320 --> 00:37:40.960
a test bed. They use the calilinux tools, They attack
558
00:37:41.039 --> 00:37:43.840
the system and demonstrate how you can get into the
559
00:37:43.880 --> 00:37:47.719
system and cause it to bring about simulated physical consequences.
560
00:37:47.719 --> 00:37:50.920
So you've taught this AI model how to use the
561
00:37:51.000 --> 00:37:54.599
calilinux tools to attack OT systems, to brick stuff and
562
00:37:54.599 --> 00:37:59.079
bring about physical consequences. You take that training model, couple
563
00:37:59.079 --> 00:38:01.159
it with the image, wrap it up in you know,
564
00:38:01.320 --> 00:38:03.880
enough code to run the image as a sort of
565
00:38:04.159 --> 00:38:08.000
kind of embedded virtual machine to run the the the
566
00:38:08.039 --> 00:38:14.559
AI model, the million by million matrix of you know numbers,
567
00:38:14.599 --> 00:38:17.760
that is a neural network. Run the neural network, run
568
00:38:17.800 --> 00:38:21.320
the KELLAI Linux image, and have the AI operate the
569
00:38:21.360 --> 00:38:24.960
tools to attack a real OT system. Drop that three
570
00:38:25.079 --> 00:38:27.239
three and a half gigabytes of attack code on an
571
00:38:27.239 --> 00:38:31.920
OT asset, start it and walk away, and it will
572
00:38:32.000 --> 00:38:34.719
figure out what's there. It will figure out how to
573
00:38:34.760 --> 00:38:36.840
attack it. It will figure out how to bring about
574
00:38:36.840 --> 00:38:42.519
physical consequences. I heard that and I thought, crap, that's nasty.
575
00:38:42.960 --> 00:38:45.840
You know, back in the day, stucksnitt was autonomous. It
576
00:38:45.960 --> 00:38:50.639
did its thing, but it was a massive investment to
577
00:38:50.920 --> 00:38:54.840
produce an asset, a piece of malware that did its
578
00:38:54.880 --> 00:39:00.440
thing without human intervention. This strikes me as again, something
579
00:39:00.440 --> 00:39:03.199
that will do its thing without human intervention and it
580
00:39:03.199 --> 00:39:05.880
will figure out as it goes. It's one
investment you
581
00:39:05.880 --> 00:39:11.880
can leverage across hundreds of different kinds of targets. I
582
00:39:11.960 --> 00:39:15.199
was alarmed. This is something I'm thinking about going forward.
583
00:39:15.960 --> 00:39:19.400
You know. It's to me, this is a credible threat.
584
00:39:19.400 --> 00:39:21.280
This is something we only need to worry about. I
585
00:39:21.280 --> 00:39:25.760
don't know that this thing exists yet, but I'm pretty
586
00:39:25.760 --> 00:39:31.679
sure it will in five years. Is everything credible? What,
587
00:39:32.239 --> 00:39:34.440
in your mind is not a credible threat?
588
00:39:34.480 --> 00:39:40.000
At this point, I would think that large scale destructive
589
00:39:40.119 --> 00:39:44.840
attacks on big machinery is not something that I would
590
00:39:44.880 --> 00:39:48.719
consider a credible attack. But it also goes back to
591
00:39:49.000 --> 00:39:52.000
the motivation of the threat actor. For instance, if you
592
00:39:52.079 --> 00:39:56.719
have a small municipality, I would see that really heavy
593
00:39:56.840 --> 00:40:01.440
sophisticated cyber attacks, a lot of them wouldn't be actually
594
00:40:01.599 --> 00:40:05.159
credible due to the target not being interesting for such
595
00:40:05.159 --> 00:40:10.880
a threat actor. So large scale destructive attacks is something
596
00:40:10.960 --> 00:40:14.920
that in a lot of scenarios wouldn't be a credible attack.
597
00:40:15.679 --> 00:40:22.639
And then we have, for instance, large scale blackout is
598
00:40:23.960 --> 00:40:26.960
quite an interesting story nowadays because a couple of weeks ago,
599
00:40:27.039 --> 00:40:30.480
I would think that it wasn't actually a credible attack.
600
00:40:31.719 --> 00:40:35.519
Once we now see that it can happen. For instance,
601
00:40:35.599 --> 00:40:39.079
with Spain, it wasn't probably not a cyber attack, but
602
00:40:39.239 --> 00:40:43.199
it was something that happened on the consequence side. If
603
00:40:43.199 --> 00:40:46.559
we can show that or identify that it actually can
604
00:40:46.599 --> 00:40:50.360
be caused by a cyber attack, then that suddenly, nowadays,
605
00:40:50.760 --> 00:40:55.559
within the last week, has become a credible attack. And
606
00:40:55.599 --> 00:41:00.960
also swarm kind of attacks. I hear discussions on that
607
00:41:01.119 --> 00:41:04.880
from time to time where they see talk about whether
608
00:41:04.920 --> 00:41:09.280
it's a credible thing where you attack at millions of cars.
609
00:41:10.760 --> 00:41:13.000
As of now, I don't see that as a credible attack,
610
00:41:13.119 --> 00:41:17.719
but things can change, you know.
611
00:41:17.760 --> 00:41:21.960
It's an interesting statement he made there that large scale
612
00:41:21.960 --> 00:41:26.519
attacks on heavy machinery isn't credible. You know, when I
613
00:41:26.559 --> 00:41:29.400
think about what we're talking about on this podcast. The
614
00:41:29.400 --> 00:41:34.360
purpose of OT security presumably is that there are significant
615
00:41:34.440 --> 00:41:39.760
risks to really important machines at large scale. But maybe
616
00:41:39.800 --> 00:41:45.159
at this point we've covered that.
617
00:41:45.199 --> 00:41:48.159
That's a good point. I think one of the lessons
618
00:41:48.159 --> 00:41:50.880
here is that determining what is and is not credible
619
00:41:51.119 --> 00:41:54.519
is a judgment call. Okay, different experts are going to disagree.
620
00:41:56.239 --> 00:41:59.360
I've you know, a few years ago, I saw research
621
00:41:59.400 --> 00:42:06.039
published saying, look, here's let's take for the sake of argument,
622
00:42:06.119 --> 00:42:08.320
the possibility of attacking a I don't know, a chemical
623
00:42:08.360 --> 00:42:12.239
plant and you know, causing a toxic discharge. And the
624
00:42:12.280 --> 00:42:16.760
researchers concluded that it was theoretically possible, but it was
625
00:42:17.119 --> 00:42:20.199
such an enormous amount of effort on the on the
626
00:42:20.199 --> 00:42:22.440
part of the adversary, all of which would have to
627
00:42:22.440 --> 00:42:24.960
go on undetected by the site. They said, you know,
628
00:42:25.039 --> 00:42:27.320
in the end, I just don't know that this is
629
00:42:27.639 --> 00:42:31.159
reasonable to believe that this will ever happen. So, you know,
630
00:42:31.199 --> 00:42:35.559
that was one site, one one data point. But again,
631
00:42:36.119 --> 00:42:39.039
you know, there are the experts. Experts disagree. This is
632
00:42:39.039 --> 00:42:42.079
the what I learned on the very first book I wrote.
633
00:42:42.159 --> 00:42:47.800
I got wildly different feedback from different internationally recognized experts.
634
00:42:49.679 --> 00:42:57.079
Here's here's an insight to me. This means that when
635
00:42:57.119 --> 00:43:01.840
we make judgments about credibility, we probably have to be
636
00:43:01.920 --> 00:43:03.480
We have to make you know, if we're going to
637
00:43:03.519 --> 00:43:05.679
make a mistake, make a mistake on the side of caution,
638
00:43:05.960 --> 00:43:10.360
error on the side of caution. Because different experts have
639
00:43:10.400 --> 00:43:14.320
different opinions, we might be wrong. You know, every expert
640
00:43:14.480 --> 00:43:17.559
has to be honest enough to admit that we might
641
00:43:17.599 --> 00:43:21.400
be wrong and build a margin for error into their
642
00:43:21.519 --> 00:43:25.880
judgment of what's credible. So even if we don't believe that,
643
00:43:26.000 --> 00:43:28.400
you know, an attack that I don't know destroys a
644
00:43:28.440 --> 00:43:32.880
turbine is credible, we might want to take some reasonable
645
00:43:33.039 --> 00:43:38.280
defenses to against you know, such a not terribly credible
646
00:43:38.320 --> 00:43:41.480
attack in our opinion, But we might want to deploy
647
00:43:41.960 --> 00:43:47.880
defenses anyway, just because we
might be wrong. And you
648
00:43:47.880 --> 00:43:50.880
know this, this is something that is also being discussed.
649
00:43:50.880 --> 00:43:53.000
It's how big a margin for error do we need
650
00:43:53.039 --> 00:43:56.840
to build into our planning. I mean, I talk to
651
00:43:56.840 --> 00:44:01.039
a gentleman who produces who designs pedestrian bris. I said,
652
00:44:01.440 --> 00:44:03.880
how do you calculate the maximum load? He says, that's easy, Andrew,
653
00:44:03.840 --> 00:44:06.400
you build a barrier to either side of the bridge.
654
00:44:07.119 --> 00:44:10.599
Vehicles can't get on the bridge. Most people are less
655
00:44:10.599 --> 00:44:12.840
than two meters tall. Most people are mostly water. You
656
00:44:12.960 --> 00:44:15.119
model two meters of water the width of the bridge,
657
00:44:15.119 --> 00:44:17.039
the length of the bridge, that's your maximum load. And
658
00:44:17.079 --> 00:44:20.199
then he says and then he says, you multiply that
659
00:44:20.239 --> 00:44:22.719
by eight, and you build the bridge to carry the
660
00:44:22.800 --> 00:44:26.679
multiplied load. Because these are people we're talking about, it
661
00:44:26.760 --> 00:44:30.960
is unacceptable for the bridge to fail under load. And
662
00:44:31.000 --> 00:44:34.360
so this is the margin for error that engineers routinely
663
00:44:34.400 --> 00:44:39.480
build into their safety calculations. I believe we as experts
664
00:44:39.480 --> 00:44:42.199
in cybersecurity need to build a margin for error into
665
00:44:42.199 --> 00:44:47.360
our security planning as well. One of the things that
666
00:44:47.760 --> 00:44:52.280
appeals to me very much about the credibility concept is
667
00:44:53.039 --> 00:44:57.599
using the concept to communicate with non technical decision makers
668
00:44:57.639 --> 00:45:01.440
like boards and directors. You do this, you have experience
669
00:45:01.480 --> 00:45:03.119
with this? Can you talk about your experience?
670
00:45:03.760 --> 00:45:06.079
Yeah. I think it's interesting when when we talk to
671
00:45:06.559 --> 00:45:13.840
talk to board members and the c xos in different companies,
672
00:45:13.880 --> 00:45:18.480
they then they don't necessarily go into details about risk,
673
00:45:18.599 --> 00:45:22.039
but they know that they have a special accountability. So
674
00:45:22.599 --> 00:45:26.639
when we talk about credibility for those kind of people,
675
00:45:26.880 --> 00:45:30.920
they are getting more on board with the discussions. They
676
00:45:30.960 --> 00:45:35.800
know they have a special accountability. They draw the line
677
00:45:35.840 --> 00:45:39.880
in the sand. For instance, if the potential consequence is
678
00:45:39.920 --> 00:45:45.480
that somebody would die, then that's a non acceptable risk
679
00:45:45.920 --> 00:45:49.519
and they take on that kind of position due to
680
00:45:49.599 --> 00:45:54.920
their accountability as board members or heads of the company.
681
00:45:55.599 --> 00:46:03.320
And they also are being accountable for from from the
682
00:46:03.320 --> 00:46:07.679
the government and from this for the society. So some
683
00:46:07.920 --> 00:46:11.239
some risks when it comes to the consequence side. If
684
00:46:11.280 --> 00:46:16.039
if we talk about people dying, then that's absolutely not
685
00:46:16.639 --> 00:46:21.400
acceptable risk for the society. And the representatives for for
686
00:46:21.800 --> 00:46:27.159
that kind of approaches is elected persons in the government
687
00:46:27.840 --> 00:46:31.719
and they put the heads of the company or the
688
00:46:31.719 --> 00:46:34.960
board of directors as accountable for that on top of
689
00:46:35.039 --> 00:46:35.519
the company.
690
00:46:36.320 --> 00:46:41.280
So that makes sense. You know, boards care about consequences
691
00:46:41.400 --> 00:46:44.719
that the business or the society is going to find unacceptable.
692
00:46:45.159 --> 00:46:47.719
You didn't use the word credible. How does credibility fit
693
00:46:47.840 --> 00:46:50.519
into acceptability when you're communicating with the board.
694
00:46:51.719 --> 00:46:55.320
Yeah, we don't have to defend against all possible cyber attacks.
695
00:46:55.360 --> 00:46:58.920
What we do have to protect against is the credible ones.
696
00:46:59.360 --> 00:47:02.480
So when we bring credibility in as a concept, then
697
00:47:02.559 --> 00:47:08.880
it's something that communicates communicates much better for the board
698
00:47:08.880 --> 00:47:11.960
of directors and the head of head of the companies.
699
00:47:13.599 --> 00:47:16.719
This has been good, but you know, it's it's a
700
00:47:16.760 --> 00:47:19.400
field big enough that I fear we've missed something. You know,
701
00:47:19.679 --> 00:47:22.119
let me ask you an open question. What should I
702
00:47:22.159 --> 00:47:22.920
have asked you? Here?
703
00:47:23.760 --> 00:47:28.320
We've been talking about credibility. Credibility is what is reasonable
704
00:47:28.400 --> 00:47:33.280
to believe. But it's not enough to talk about reasonable attacks.
705
00:47:33.400 --> 00:47:37.760
We also need to be talking about reasonable defense. So
706
00:47:37.840 --> 00:47:40.920
what is a reasonable defense. We then need to be
707
00:47:41.320 --> 00:47:44.360
considering or taking all the tools.
708
00:47:45.599 --> 00:47:46.519
We need to use, all the.
709
00:47:46.480 --> 00:47:50.920
Tools at our disposal for a reasonable defense, and nowadays
710
00:47:51.400 --> 00:47:55.719
that also obviously includes AI on the defensive side, not
711
00:47:55.760 --> 00:47:59.440
only on the offensive side. This is also a very
712
00:47:59.440 --> 00:48:02.480
important part part of me of the reason for me
713
00:48:02.760 --> 00:48:08.280
joining Omni. So Omni is built on our security knowledge graph,
714
00:48:08.559 --> 00:48:12.039
So it's a data model where we can put all
715
00:48:12.079 -->
00:48:15.800
information we need about our assets, on the vulnerabilities, on
716
00:48:15.840 --> 00:48:19.880
the network topologies, on the threats, the threat actors. So
717
00:48:19.920 --> 00:48:23.039
it becomes a digital representation or a digital twin of
718
00:48:23.119 --> 00:48:27.079
our assets. Combining that with AI, which we have built
719
00:48:27.079 --> 00:48:30.960
in from the beginning, we get a very strong assistant
720
00:48:31.079 --> 00:48:34.599
on security where it matters most.
721
00:48:35.639 --> 00:48:38.079
This has been great. Thank you Kenneth for joining us
722
00:48:38.840 --> 00:48:40.519
before I let you go. Can I ask you to
723
00:48:40.599 --> 00:48:43.119
sum up for our listeners? What should we take away
724
00:48:43.119 --> 00:48:44.199
from this episode?
725
00:48:44.960 --> 00:48:47.280
Thank you Andrew for having me, and thank you so
726
00:48:47.400 --> 00:48:50.519
much for being here in Norway and visiting us at
727
00:48:50.559 --> 00:48:55.079
our office. So we've had a good conversation about consequence,
728
00:48:55.599 --> 00:49:01.440
the focus on the worst case consequences, where we moved
729
00:49:01.440 --> 00:49:09.639
over to talking about credibility, replacing the likelihood concept with credibility,
730
00:49:09.800 --> 00:49:14.320
especially for high impact stuff where we don't have the
731
00:49:14.360 --> 00:49:19.400
probability or the data to talk about it. We also
732
00:49:19.440 --> 00:49:28.000
talked about reasonable attacks and reasonable defenses, So what is
733
00:49:28.039 --> 00:49:38.800
a reasonable defense against increasingly credible, sophisticated attacks with high consequences.
734
00:49:39.360 --> 00:49:43.559
So it's been a really good discussion about all of
735
00:49:43.599 --> 00:49:48.119
these topics. If people wanted to know more about these
736
00:49:48.199 --> 00:49:55.239
topics or they want to discuss them, please connect with
737
00:49:55.320 --> 00:49:59.000
me on LinkedIn Message me there. I'm more than a
738
00:49:59.000 --> 00:50:03.280
happy discussion to discuss these topics. Please visit our webpage
739
00:50:03.800 --> 00:50:08.480
omnisecurity dot com. Our platform addresses most of these topics
740
00:50:08.519 --> 00:50:09.719
we talked about today.
741
00:50:13.159 --> 00:50:15.880
Andrew, that just about does it for your conversation with
742
00:50:16.079 --> 00:50:19.360
Kenneth Tittelstadt. Do you have any final words you would
743
00:50:19.360 --> 00:50:21.920
like to take out our episode with today.
744
00:50:22.840 --> 00:50:25.760
Yeah, I mean we've we've talked about about credibility, and
745
00:50:25.880 --> 00:50:28.599
this is a concept that is relevant to sort of
746
00:50:28.599 --> 00:50:31.840
the high end of sophisticated attacks, the high end of
747
00:50:32.840 --> 00:50:39.000
of consequence. But you know, I'm not sure. Let me
748
00:50:39.159 --> 00:50:41.559
let me try and and and give a very simple example.
749
00:50:41.599 --> 00:50:44.159
I mean, I was I was raised in Brooks, alberta
750
00:50:45.119 --> 00:50:47.639
little town, you know, ten thousand people in the middle
751
00:50:47.679 --> 00:50:51.599
of nowhere, literally an hour's drive from any larger population center.
752
00:50:53.440 --> 00:50:56.079
You know, in terms of cyber threats, do let's pick
753
00:50:56.159 --> 00:50:58.159
let's pick on I don't know the Russian military. Does
754
00:50:58.159 --> 00:51:02.880
the Russian military have the money to buy three absolute
755
00:51:02.960 --> 00:51:06.840
cyber gurus, train them up on water systems, plant them
756
00:51:06.880 --> 00:51:10.000
as a sleeper cell in the workforce of the town
757
00:51:10.039 --> 00:51:12.719
of Brooks water treatment system, have them sit on their
758
00:51:12.719 --> 00:51:15.920
hands for three years, and after three years, using the
759
00:51:15.920 --> 00:51:19.199
passwords they've gained, the trust they've gained and the expertise
760
00:51:19.239 --> 00:51:22.280
that they have, have them launch a crippling cyber attack
761
00:51:22.360 --> 00:51:25.039
that the damages equipment, that takes the water treatment system
762
00:51:25.079 --> 00:51:29.760
down for forty five days. Is that a credible threat? Well,
763
00:51:29.840 --> 00:51:33.000
the Russians have the money to do that it's you know,
764
00:51:33.039 --> 00:51:36.400
they have the capability to do that, but you have
765
00:51:36.480 --> 00:51:39.719
to ask, why would they bother? I mean, this is
766
00:51:39.760 --> 00:51:42.280
a little agricultural community, there's a little bit of oil
767
00:51:42.280 --> 00:51:46.599
and gas activity. Why would they bother? That does not
768
00:51:46.800 --> 00:51:49.480
seem to be It does not seem to be reasonable
769
00:51:49.519 --> 00:51:52.679
to launch that kind of attack against the town of Brooks.
770
00:51:52.960 --> 00:51:55.400
It just makes no sense. I don't see that as
771
00:51:55.440 --> 00:51:59.320
a credible threat. Is that a credible threat for the
772
00:51:59.360 --> 00:52:02.400
water treatment system in the city of Washington, d C?
773
00:52:02.920 --> 00:52:06.840
Home of the Pentagon. I do think that's a credible threat.
774
00:52:06.920 --> 00:52:10.039
So the question of what's credible is an important question
775
00:52:10.159 --> 00:52:12.880
that I see one and more people asking in risk
776
00:52:12.880 --> 00:52:15.960
analysis going forward. You know, we have to figure out
777
00:52:15.960 --> 00:52:20.599
what's credible for us? You know, are what capabilities do
778
00:52:20.679 --> 00:52:24.480
our adversaries have? What kind of assets are we protecting,
779
00:52:24.480 --> 00:52:27.320
what kind of defenses do we have deployed? What makes sense?
780
00:52:27.360 --> 00:52:30.199
What's reasonable to believe in terms of the bad guys
781
00:52:30.280 --> 00:52:33.119
coming after us? This is an important question going forward,
782
00:52:33.440 --> 00:52:36.400
and I see lots of people discussing it. I'm grateful
783
00:52:36.440 --> 00:52:39.800
for the chance to explore the concept here with with Kenneth.
784
00:52:40.440 --> 00:52:43.159
Well. Thanks to Kenneth for exploring this with us, and
785
00:52:43.159 --> 00:52:45.360
Andrew is always thank you for speaking with me.
786
00:52:46.039 --> 00:52:47.320
It's always a pleasure. Thank you, Nick.
787
00:52:48.360 --> 00:52:52.480
This has been the Industrial Security podcast from Waterfall. Thanks
788
00:52:52.480 --> 00:53:00.760
to everyone out there listening inst