1
00:00:00,000 --> 00:00:05,960
well looked like coolest one of the most
things they needed a surname so I just

2
00:00:05,960 --> 00:00:14,489
in time to talk a lot actually stands
for the arithmetic unit control so it's

3
00:00:14,490 --> 00:00:20,930
nice to be here and see all your ladies
and gentlemen it's it's awesome to be on

4
00:00:20,930 --> 00:00:24,410
this conference and I want to just think
those killers of trainings for

5
00:00:24,410 --> 00:00:28,779
organizing this great event it's a big
honor to be here

6
00:00:28,779 --> 00:00:33,380
tributary we want to talk about pgp
actually we don't want to talk about PHP

7
00:00:33,380 --> 00:00:41,710
and particular bunch more about what is
broken publicly crypto this lecture goes

8
00:00:41,710 --> 00:00:47,250
and several parts so in the first part
of the lecture I decided to make sure we

9
00:00:47,250 --> 00:00:53,469
all speak the same language so we have a
little recap of what are so here's just

10
00:00:53,469 --> 00:00:55,600
a lot of stuff like this

11
00:00:55,600 --> 00:01:01,520
the second part we gonna talk about the
fun stuff how to a tacit and in the

12
00:01:01,520 --> 00:01:04,470
third part we're going to talk about

13
00:01:04,470 --> 00:01:15,479
model of trust we live in the moment so
and I tolerant of my job as break things

14
00:01:15,479 --> 00:01:22,780
so I'm hacker sure always terrible
always drinking and still learning as

15
00:01:22,780 --> 00:01:29,220
all of us there are some little
attention for the stalk don't worry it's

16
00:01:29,220 --> 00:01:34,460
not gonna be heavy on math but few
formulas in the talk so we have to

17
00:01:34,460 --> 00:01:39,649
recognize notations which all of you
know so explain to us why the ciphertext

18
00:01:39,650 --> 00:01:43,670
Easter encryption function digital
equipment function and we have 2k the

19
00:01:43,670 --> 00:01:47,500
key and sure enough a case-based but I
don't think we need to send this letter

20
00:01:47,500 --> 00:01:55,200
the questions while we do that well
that's the reason why actually it's a

21
00:01:55,200 --> 00:02:00,320
funny pictures you know that as tus
embassy in Berlin and actually project

22
00:02:00,320 --> 00:02:03,240
at on the embassy

23
00:02:03,240 --> 00:02:08,579
Yahoo last year actually order year
before last year Mr Snowdon decided to

24
00:02:08,580 --> 00:02:13,420
have a little talk with the public and
the public found out what we don't never

25
00:02:13,420 --> 00:02:21,170
knew that secret services doing Secret
Service and spice are spying not just in

26
00:02:21,170 --> 00:02:27,000
the movie's producers because I get paid
for it and you know who's gonna pay them

27
00:02:27,000 --> 00:02:34,200
us and so we now decide that's not cool
and the question is what can we do

28
00:02:34,200 --> 00:02:39,619
against it and I always see rule to
encrypt your message that secure so well

29
00:02:39,620 --> 00:02:47,080
we do and Krypton matches message and
see how secure others who have you ever

30
00:02:47,080 --> 00:02:54,220
played for public key crypto and the
rest of us on HTTPS site

31
00:02:54,220 --> 00:02:57,440
well

32
00:02:57,440 --> 00:03:04,799
founding paper public equipped was done
by actually treat people shouldn't have

33
00:03:04,799 --> 00:03:10,549
to feel Helmand and Merkel they had two
founding paper which was actually

34
00:03:10,549 --> 00:03:17,640
published 1976 and was probably the
first scheme which was introduced of

35
00:03:17,640 --> 00:03:24,380
public-key cryptography so what we do is
we have some public power made us we

36
00:03:24,380 --> 00:03:27,820
take a journey group I G

37
00:03:27,820 --> 00:03:35,760
we have some pictures which we might do
some prime soil or whatever and then we

38
00:03:35,760 --> 00:03:44,048
have some generator of the group and
what happens this we exchange keys Alice

39
00:03:44,049 --> 00:03:49,380
answer key or her message eg today a
race to the g2 race neutral palette hey

40
00:03:49,380 --> 00:03:55,829
Mike P what does the same before he
departs be than both of them gonna

41
00:03:55,829 --> 00:04:02,160
compute their private message and what
kind of happiness we have two problem to

42
00:04:02,160 --> 00:04:10,489
compute G today be out of GTA and GDP
from public information that is now

43
00:04:10,489 --> 00:04:15,700
since 1976 the case and we still have
two underlying

44
00:04:15,700 --> 00:04:21,829
other ambitious discrete logarithm and
we still think these problems heart we

45
00:04:21,829 --> 00:04:28,410
use our assay who have a used car sales
you guys so you know how it works

46
00:04:28,410 --> 00:04:35,070
shared key public key and you have a
private key and those kids are actually

47
00:04:35,070 --> 00:04:42,290
a product of a of a of a NES the
multiplication of people like you more

48
00:04:42,290 --> 00:04:47,050
or less the encryption exponent all the
decryption exponent so we exchange our

49
00:04:47,050 --> 00:04:52,710
keys to send a message and other one can
decrypt to keep the stalkers not about

50
00:04:52,710 --> 00:04:59,888
how r say works as just a quick overview
I expect you actually to no doubt so

51
00:04:59,889 --> 00:05:08,810
given that and see compute sold out so i
computer attacks attacks to the rest of

52
00:05:08,810 --> 00:05:13,750
the politics opponent as the same like
seem odd and is a problem which we think

53
00:05:13,750 --> 00:05:20,000
as heart if you have to find out of
public information so we have a math

54
00:05:20,000 --> 00:05:26,410
program which caused sage who know sage
such as awesome so we can do to sinn

55
00:05:26,410 --> 00:05:32,400
sage and that's actually how it looks
like so we have here a random crime

56
00:05:32,400 --> 00:05:40,349
which we have to keep our 512 the
computer to raise the two actually most

57
00:05:40,350 --> 00:05:46,860
used exponent and the internet and we
encrypt a message from Monday's equation

58
00:05:46,860 --> 00:05:55,900
is wrong what is it does anybody know
what is the status bar say but what is

59
00:05:55,900 --> 00:06:04,159
wrong with it which is used it the way
it is okay let's raise the to the power

60
00:06:04,160 --> 00:06:15,009
of 2024 which is used to waitress why
not have no idea what

61
00:06:15,009 --> 00:06:29,310
why

62
00:06:29,310 --> 00:06:33,570
I'm not talking about us i'm talking
about the bigger problem we have no

63
00:06:33,570 --> 00:06:41,480
padding and the seclusion don't want to
do this and and and and real world there

64
00:06:41,480 --> 00:06:54,350
is no padding so what we do there as we
factor actually are primes and reduces

65
00:06:54,350 --> 00:07:00,280
the attacks I i talked up may be reduced
to factorization maybe not all factors

66
00:07:00,280 --> 00:07:04,049
should not be np-complete on unless
something really strange happens and we

67
00:07:04,050 --> 00:07:11,070
just ran into it but now we got to the
Real Talk the question is what is the

68
00:07:11,070 --> 00:07:17,659
attack types what do you think how we
attacked us so we have a public key

69
00:07:17,660 --> 00:07:26,490
crypto and we supposed to be the
attacker what could we do which social

70
00:07:26,490 --> 00:07:36,960
engineering yes this is all what else
could we do what

71
00:07:36,960 --> 00:07:43,630
what is mine and the Middle how can you
man in the middle of one day have

72
00:07:43,630 --> 00:07:50,139
encrypted channel you have to fast break
down quickly before it can really man in

73
00:07:50,139 --> 00:07:55,800
the middle or you do an attack which are
mentioned much later in this talk but

74
00:07:55,800 --> 00:08:00,860
what else do we have no more through
theoretical models of attack what could

75
00:08:00,860 --> 00:08:15,590
we do he said social engineering very
right I want one

76
00:08:15,590 --> 00:08:28,039
planning implementation we have to
implementation floor and we have to

77
00:08:28,039 --> 00:08:34,098
social engineering is there anything
else we could use we could start trying

78
00:08:34,099 --> 00:08:46,279
to steal the private key which social
engineering if we still got that has

79
00:08:46,279 --> 00:08:50,279
actually social engineering because he
go to his place and you attack the

80
00:08:50,279 --> 00:08:54,040
machine directly so you have to go to
the machine try to go to my machine

81
00:08:54,040 --> 00:08:58,709
without my permission to have to sort in
January according to your way to the

82
00:08:58,709 --> 00:09:05,060
machine down more attacks then we have
proved false as an attack would you

83
00:09:05,060 --> 00:09:10,780
agree no yes no yes

84
00:09:10,780 --> 00:09:17,630
and we can do another think we can
analyze as an attack so we just gather a

85
00:09:17,630 --> 00:09:23,780
lot of staff and we analyze what we find
and that's actually what we're doing so

86
00:09:23,780 --> 00:09:27,560
I talked two types is proved false
analyzing social engineering and using

87
00:09:27,560 --> 00:09:36,979
implementation floss right clinton
metaxas very funny but the moment we're

88
00:09:36,980 --> 00:09:42,440
not as ready yet but we're getting there
don't worry so which type of attack do

89
00:09:42,440 --> 00:09:49,350
we have for sure enough the most juice
22 attack are safe for example is

90
00:09:49,350 --> 00:09:55,980
factoring so there are different ways of
factoring it so you have some some other

91
00:09:55,980 --> 00:10:02,090
reason which are used for specific cases
you have some random Algeria so you know

92
00:10:02,090 --> 00:10:06,720
that our algorithm like Paula troll
poverty minus one serving I think the

93
00:10:06,720 --> 00:10:13,650
most used in the moment as number fill
serving papers published so we don't

94
00:10:13,650 --> 00:10:18,740
need to go into that that is what we in
the moment currently used to brute force

95
00:10:18,740 --> 00:10:24,480
does things we use number field sieve
into factor to find common factor which

96
00:10:24,480 --> 00:10:30,820
we maybe can divide to calculator
computes the private key but there was a

97
00:10:30,820 --> 00:10:40,930
paper which came out I think 2013 from
last page I think a computer too much

98
00:10:40,930 --> 00:10:47,739
does it cost to factor and if you could
look at it like the 512 bits is quite

99
00:10:47,740 --> 00:10:54,370
easy going over 20 becomes more
expensive but it still is still doable

100
00:10:54,370 --> 00:11:01,340
and here we go out thing tolerant on
three-letter service is capable of doing

101
00:11:01,340 --> 00:11:07,330
so I think they get the money are under
review time which means it's a

102
00:11:07,330 --> 00:11:13,410
non-muslim ranting on Amazon we don't
talk about their own networks I mean if

103
00:11:13,410 --> 00:11:17,900
you look at Anastasia piles of computers
are thinking of like hold stocks of all

104
00:11:17,900 --> 00:11:20,250
the computers so

105
00:11:20,250 --> 00:11:28,790
they they have enough power to compute
that right but if we look at this maybe

106
00:11:28,790 --> 00:11:38,319
it's too much right yeah so that means
we just use a large key and we say but

107
00:11:38,320 --> 00:11:47,570
sure we're safe so that takes out a
brute-force attack but we still can

108
00:11:47,570 --> 00:11:54,460
analyze stuff to social engineering and
juice implementation floss let's go for

109
00:11:54,460 --> 00:11:59,700
that task will look like in the Praxis
what what we gonna have what we find on

110
00:11:59,700 --> 00:12:04,380
the wild to be really find the perfect
round the last we will find everything

111
00:12:04,380 --> 00:12:11,900
perfect and place we have find a lot of
random generation plus we find repeat

112
00:12:11,900 --> 00:12:16,870
his keys and key collisions we find or
actually the we don't find but we have a

113
00:12:16,870 --> 00:12:26,839
very funny trust model so we do trust
people we do not know for any reason our

114
00:12:26,839 --> 00:12:36,050
secret keys so random number generation
flaws what can we do we can collect a

115
00:12:36,050 --> 00:12:41,979
lot of taters which you find on the
internet you find keys you find Cyprus

116
00:12:41,980 --> 00:12:46,500
you find websites with their search to
find all kinds of stuff on the internet

117
00:12:46,500 --> 00:12:55,870
and if you want to shop for it we just
go to her and find everything want to

118
00:12:55,870 --> 00:13:02,970
sign and cook two sessions signs we try
to cyber loss that's a password is not

119
00:13:02,970 --> 00:13:06,970
cool site in my conference when you say
site where you have to drink as you get

120
00:13:06,970 --> 00:13:08,230
my stuff

121
00:13:08,230 --> 00:13:15,790
so i'm good for twins rings already so
you can stop what you find everywhere

122
00:13:15,790 --> 00:13:21,650
for the question is how do we do that we
can actually go back to projects that

123
00:13:21,650 --> 00:13:27,650
did it already so if we don't wanna work
we can be lazy and we can use the eff

124
00:13:27,650 --> 00:13:34,819
you can download these boxes but the
results are not new I think the results

125
00:13:34,820 --> 00:13:40,340
of the 2010-12 something like this but
most of the service still use the same

126
00:13:40,340 --> 00:13:46,370
key so that's not really a big problem
we could sure enough to get us 23 ec2

127
00:13:46,370 --> 00:13:52,550
instances and run them through and not
the internet that takes a while we can

128
00:13:52,550 --> 00:14:01,680
use disposable internet study in 2008
but I think that's kind of all too soon

129
00:14:01,680 --> 00:14:05,489
after hackers of you can use a botnet
you just wait till dollar and important

130
00:14:05,490 --> 00:14:10,930
attacks are not working take it over and
use this for and we're eating what we

131
00:14:10,930 --> 00:14:15,359
can do in the US much faster as anybody
played with you from you guys played

132
00:14:15,360 --> 00:14:22,770
this season I already did you like it
they have some flaws in the beginning I

133
00:14:22,770 --> 00:14:23,480
mean

134
00:14:23,480 --> 00:14:28,170
attempt to crush always feel session
widened and the moment implementation I

135
00:14:28,170 --> 00:14:34,170
worked quite well so I tried to some you
don't like it

136
00:14:34,170 --> 00:14:41,269
that's all that's awesome I mean I use
it alot the last time i used i didnt

137
00:14:41,269 --> 00:14:46,339
juicy tomato internet but I was I try to
sleep it wasn't talking in a hotel I try

138
00:14:46,339 --> 00:14:49,970
to sleep in there was a lady on the
floor and she was skyping all the time

139
00:14:49,970 --> 00:15:01,990
and I could not sleep so what I did was
to us like instantly amazing thing about

140
00:15:01,990 --> 00:15:07,300
it even in the plaza hotel in 10 hours I
had like the whole internet so I told

141
00:15:07,300 --> 00:15:12,910
goal of us like I think twenty to go
back to port 443 to make sure I can

142
00:15:12,910 --> 00:15:14,790
sleep

143
00:15:14,790 --> 00:15:21,810
the question is what do we find one we
when we play a round over 200 + find

144
00:15:21,810 --> 00:15:28,138
like 30 million open ports I didn't 2013
maybe we need to update the numbers

145
00:15:28,139 --> 00:15:34,389
right so just numbers are a little bit
older we have had like a six million or

146
00:15:34,389 --> 00:15:45,350
say Keyes you'll find about 6,000 to use
a 10 and two hundred kids are you as you

147
00:15:45,350 --> 00:15:47,339
know I know you need to know

148
00:15:47,339 --> 00:15:57,279
Russian Standard why we are on Russian
Standard vodka that's my problem I'm my

149
00:15:57,279 --> 00:16:01,370
most favorite brand of Russian Standard
yes it's a russian standards do you know

150
00:16:01,370 --> 00:16:10,420
when they introduced it no I mean when
they introduced it actually it was

151
00:16:10,420 --> 00:16:14,189
introduced and that's funny because I'm
voting numbers of its John McCarthy

152
00:16:14,190 --> 00:16:23,730
always like this was introduced on May
20 30 1994 so we have a 323 as you could

153
00:16:23,730 --> 00:16:26,970
see before I that's why I always too
much computation study have a

154
00:16:26,970 --> 00:16:28,870
twenty-three in there

155
00:16:28,870 --> 00:16:55,740
yes yes sure what do we find we find SSH
keys we find PGP GPG stuff if I get

156
00:16:55,740 --> 00:17:08,069
caught and there are some national issue
documents as a result of yourself what

157
00:17:08,069 --> 00:17:11,230
you can do it on white year because they
were questioned that walk when everybody

158
00:17:11,230 --> 00:17:15,750
try to do that maybe you want to do with
the university or at home a few other

159
00:17:15,750 --> 00:17:20,900
bigger lying 1 gigabit line it should be
done in a few hours so that's that's ok

160
00:17:20,900 --> 00:17:26,959
so what you find on as ages before we
had to say now we have to as a church we

161
00:17:26,959 --> 00:17:32,980
have 25 million open ports which we fall
million hace keys to us numbers you can

162
00:17:32,980 --> 00:17:40,200
read it by yourself that's not really
important for the talk so the question

163
00:17:40,200 --> 00:17:48,920
is how do we find we want to we want we
want to play with PGP and we sat as as

164
00:17:48,920 --> 00:17:54,530
an attack type analytics would be ok
then we say so so how do we analyze the

165
00:17:54,530 --> 00:18:03,290
PGP what's about doing your own case
over that's nadia is not so you start

166
00:18:03,290 --> 00:18:08,139
your own ski service just analyze what's
on it to start your case over you just

167
00:18:08,140 --> 00:18:14,670
downloaded and then you can look into it
what you find in this case over numbers

168
00:18:14,670 --> 00:18:21,410
are down here at such a place over 2013
last two years were too busy to to play

169
00:18:21,410 --> 00:18:27,230
that we had a talk about Bitcoin ten
o'clock today I think there was two

170
00:18:27,230 --> 00:18:29,799
awesome people talking about Bitcoin
right

171
00:18:29,799 --> 00:18:39,369
so it can uses electric elliptical to
feel at ease while she is a kind of

172
00:18:39,369 --> 00:18:43,539
security and you have a good ass but if
you notice fucked up you're not really

173
00:18:43,539 --> 00:18:49,350
secure and trust me the gnosis really
fucked up the most people since taking

174
00:18:49,350 --> 00:18:56,830
some snake also produces so if you want
to get a stick on change you just open

175
00:18:56,830 --> 00:19:00,110
yourself a pic on account and you
download to change the locks changed

176
00:19:00,110 --> 00:19:06,330
there and then we can start on player
but you will have a download which is

177
00:19:06,330 --> 00:19:18,989
found about 76 and that gives you one
million 870,000 keys roundabout about 24

178
00:19:18,989 --> 00:19:29,019
million transactions there was to
research from the guise of Palestine and

179
00:19:29,019 --> 00:19:33,749
a few other guys and they actually made
a paper and what it did

180
00:19:33,749 --> 00:19:39,799
taiwan they had some government issued
ID cards and you could do transactions

181
00:19:39,799 --> 00:19:44,619
with them and they had a trip on them
and they actually brokered to find the

182
00:19:44,619 --> 00:19:50,820
paper we found the paper right that's
awesome paper in his readers are you

183
00:19:50,820 --> 00:19:54,029
guys should read this paper that's
awesome

184
00:19:54,029 --> 00:19:58,059
flights will be online soon you just
copy pasted all of the slights and you

185
00:19:58,059 --> 00:20:03,940
can download it interpreted it as they
actually took over identities they could

186
00:20:03,940 --> 00:20:08,419
do transactions so we really need to
think about what we doing because that

187
00:20:08,419 --> 00:20:14,379
now goes into our private sector and
into our entire families when we have

188
00:20:14,379 --> 00:20:23,918
IDs installing what we look for sure
enough we look for repeated case we look

189
00:20:23,919 --> 00:20:30,259
on public which are not really got well
made and we look for a four-shot factors

190
00:20:30,259 --> 00:20:35,309
so sharp factors the formula would say
would look like this

191
00:20:35,309 --> 00:20:40,269
because those guys who have shared
factors use the same product key so

192
00:20:40,269 --> 00:20:42,620
that's what we're looking for

193
00:20:42,620 --> 00:20:50,389
so what can we do whatever we could
either fraud are instances and trial

194
00:20:50,389 --> 00:20:58,370
number for cheering serving on it if the
key 768 bet he would take 2.5 calendar

195
00:20:58,370 --> 00:21:08,518
years on ec2 units which use but if you
used to achieve city but if we take the

196
00:21:08,519 --> 00:21:23,200
same key almost 50 90 seconds I talk
talk about all around them laptop so ok

197
00:21:23,200 --> 00:21:31,289
that's ok I'm choosing to you say anyway
fact that we paid off of DSA not really

198
00:21:31,289 --> 00:21:38,440
so if we look at it and yes I is too
long skirts known the private key is

199
00:21:38,440 --> 00:21:45,129
computable so in the beginning on the
same like on the beginning of a few case

200
00:21:45,129 --> 00:21:50,389
you need something to be done and and he
has ever started in on sunday scapes

201
00:21:50,389 --> 00:21:52,060
known

202
00:21:52,060 --> 00:21:56,740
we can compute its over if you find two
years the same as such that the same

203
00:21:56,740 --> 00:22:07,010
Lance case easy to compute but case
unknown right if we have to

204
00:22:07,010 --> 00:22:12,870
message to actually who share the same
los que que Cambie computers from the

205
00:22:12,870 --> 00:22:17,110
signature of those two messages so one
way to reduce with obstructive

206
00:22:17,110 --> 00:22:24,290
signatures South K and worse because
that's all you ok can be computed

207
00:22:24,290 --> 00:22:36,510
private keys then who knows about the
paradox are right for us who don't know

208
00:22:36,510 --> 00:22:40,470
about the paradox of osteoporosis
protect my which says if you want to

209
00:22:40,470 --> 00:22:47,260
compute if we have 23 people in a room
the chances are fifty 50 that we have a

210
00:22:47,260 --> 00:22:54,740
coalition right we have we have a higher
chance of 99.2% as people have a

211
00:22:54,740 --> 00:23:00,610
different birthday but we have already a
50% collision chance that means that we

212
00:23:00,610 --> 00:23:05,250
have a coalition that's actually what
we're looking for in the end

213
00:23:05,250 --> 00:23:10,200
crypto crypto if you too crypto analyze
the script analysis of most of the time

214
00:23:10,200 --> 00:23:15,500
it's guessing looking for collisions
looking for repetition so that we find

215
00:23:15,500 --> 00:23:22,020
repeated cason stuff that's the way you
which are probably due to sinn sage if

216
00:23:22,020 --> 00:23:31,800
you wanna play but so why do we talk
about it because the randomness we maybe

217
00:23:31,800 --> 00:23:37,040
have a floor some random this there was
this paper fact about natural strict

218
00:23:37,040 --> 00:23:45,310
about that so you're in the crypto and
you definitely can tell so that you can

219
00:23:45,310 --> 00:23:51,070
find all kinds of stuff and Bernstein
again he had actually this little theory

220
00:23:51,070 --> 00:23:59,770
where he say we stopped and ate our same
module and 1924 5 and so on

221
00:23:59,770 --> 00:24:05,730
and computer product of this module loss
on a binary tree and we model the

222
00:24:05,730 --> 00:24:12,860
product by the Square after containing
box containing notes and then our goal

223
00:24:12,860 --> 00:24:19,780
is to go and 12 13 14 March and to the
power to about and want to power to and

224
00:24:19,780 --> 00:24:26,730
so on and divider to fire find purchase
city some non-trivial devised advisers

225
00:24:26,730 --> 00:24:31,890
that's what we want to find out if you
find some non-trivial devices by using

226
00:24:31,890 --> 00:24:37,740
the GCT if you go by the results we
found you'll find out that you have to

227
00:24:37,740 --> 00:24:43,580
stay city maybe in ten hours if you walk
this way undermine Eritrea it's it's

228
00:24:43,580 --> 00:24:48,240
well written on the website did you find
it while you're looking on at all anyway

229
00:24:48,240 --> 00:24:50,760
did you find this article from him

230
00:24:50,760 --> 00:24:57,570
you should check our chats about the
binary tree sits very awesome because

231
00:24:57,570 --> 00:25:04,530
what we do right now is more statistics
10 attack type b2 analyst we analyze and

232
00:25:04,530 --> 00:25:09,149
what we do here is a statistic way of
analyzing our factoring case instead of

233
00:25:09,150 --> 00:25:17,990
doing the hard way and you should be
using our cities so if you look at our

234
00:25:17,990 --> 00:25:24,330
results will find about and now we're
talking about private case we talk about

235
00:25:24,330 --> 00:25:31,629
privacy case from 64,000 HP servers so
we talking about here

236
00:25:31,630 --> 00:25:36,780
private case once you have privately
assured of can decrypt stuff but you

237
00:25:36,780 --> 00:25:42,580
also can look for their secrets and the
key because of what I found was that guy

238
00:25:42,580 --> 00:25:46,050
who had a really weak private key

239
00:25:46,050 --> 00:25:51,560
US passports and and that of a day his
new key is very strong but the same

240
00:25:51,560 --> 00:26:04,310
password they don't help the same dose
of the arse des annonces you have to do

241
00:26:04,310 --> 00:26:12,260
you have a lot of the HD's a service who
has a coalition announces a profound 158

242
00:26:12,260 --> 00:26:20,300
conferences which too much about money
so why is that with the TSA you have to

243
00:26:20,300 --> 00:26:29,050
remember if a key is not well created
most of the machines have both so you

244
00:26:29,050 --> 00:26:33,820
can figure if the US is fucked up the
ass fucked up to

245
00:26:33,820 --> 00:26:41,790
so we have some repeated caisson as they
chose so why do we have repeatedly sure

246
00:26:41,790 --> 00:26:47,010
enough to companies who issue 1 keep all
their websites and you have like virtual

247
00:26:47,010 --> 00:26:51,970
hosting all kinds of stuff corporate so
its default device with the fall case

248
00:26:51,970 --> 00:26:59,370
but well

249
00:26:59,370 --> 00:27:08,520
where do we have to most problems the
most problems so called industrial grade

250
00:27:08,520 --> 00:27:15,860
firewalls we're talking about the stuff
from Cisco John apart from Sophos

251
00:27:15,860 --> 00:27:22,179
machines stuff like this all we do have
actually a problem are embedded devices

252
00:27:22,180 --> 00:27:28,300
I did the research with heavy and
pittsburgh has anybody knows this box

253
00:27:28,300 --> 00:27:32,840
it's very very common in Germany and
that's a little router which connected

254
00:27:32,840 --> 00:27:34,760
to the Internet

255
00:27:34,760 --> 00:27:42,309
they really suck so we have to pay
issues you know how the centerpiece

256
00:27:42,309 --> 00:27:47,960
generator who doesn't know everybody
know how to answer beyond your limits

257
00:27:47,960 --> 00:27:55,250
devices generated right at least you
think you don't know how right you know

258
00:27:55,250 --> 00:27:59,670
at least you think you know because
there's some surprises normally it

259
00:27:59,670 --> 00:28:04,210
should be like this you have two tough
random and that takes some key what time

260
00:28:04,210 --> 00:28:09,420
in mouse movement network traffic
whatever rights and we have to touch you

261
00:28:09,420 --> 00:28:16,120
random one is a good website another
number generator one is a good one do

262
00:28:16,120 --> 00:28:28,309
you want us one is that actually that
you want to have that thing yet but

263
00:28:28,309 --> 00:28:32,250
there's actually a real world
application would I say it have to be

264
00:28:32,250 --> 00:28:37,970
there for security I'm not talking about
for not security i'm talking bout for

265
00:28:37,970 --> 00:28:41,610
security and real world application
where we need to absolute number

266
00:28:41,610 --> 00:28:44,240
generator for security

267
00:28:44,240 --> 00:28:49,460
do you think you know which one is this
what could it be

268
00:28:49,460 --> 00:28:58,500
what for passing what else padding for
what else could we use it

269
00:28:58,500 --> 00:29:06,539
everybody's of you guys using a
telephone to use of one part time and

270
00:29:06,539 --> 00:29:13,899
what it do you go i play now you go from
your little cell phone to at our cat

271
00:29:13,899 --> 00:29:19,449
decrypt it get an encrypted to the other
tower and get back and put it in other

272
00:29:19,450 --> 00:29:24,890
cell phone and what happens on this
encryption has to sort out using a

273
00:29:24,890 --> 00:29:28,179
one-time pad and they have a
pseudorandom generator because they need

274
00:29:28,179 --> 00:29:32,750
to generate the same number on both
sides understanding of receiving such if

275
00:29:32,750 --> 00:29:38,210
you have a true random generator that
that would be paying so that's that's a

276
00:29:38,210 --> 00:29:42,580
useful way of using absolute are now
running on generator but for everything

277
00:29:42,580 --> 00:29:44,610
else we might not want to do it

278
00:29:44,610 --> 00:29:50,178
the problem of the random generator is
blocking behavior so just figure you

279
00:29:50,179 --> 00:29:57,710
have dollar and a little rude approved
device and the three wise wanna boot up

280
00:29:57,710 --> 00:30:01,399
or not boot up because it doesn't have a
random generator because i dont have

281
00:30:01,399 --> 00:30:08,580
entropy so want to put up so to
generator to need some software like

282
00:30:08,580 --> 00:30:15,629
OpenSSL liberals or whatever you prefer
to use right but the problem is not a

283
00:30:15,630 --> 00:30:20,059
set of the random last saw your device
don't have anything to seat randomness

284
00:30:20,059 --> 00:30:25,379
in the use the default an open-access
which is who knows it

285
00:30:25,380 --> 00:30:33,029
what is a smart people in in actually
implement their to use our instead of

286
00:30:33,029 --> 00:30:41,789
having like real seat what to use the
time which is awesome so this is

287
00:30:41,789 --> 00:30:47,279
actually a coat here for some people who
wear doesn't boot up and the little the

288
00:30:47,279 --> 00:30:54,759
router devices so we used to time while
the time timing problem

289
00:30:54,759 --> 00:31:02,749
you know which time they use a lot of
time Unix time so what about we have a

290
00:31:02,749 --> 00:31:07,419
lot of little fits boxes using all unix
time are we switching on the same time

291
00:31:07,419 --> 00:31:18,229
so if you remember we have we created
our psyche if we have too little for its

292
00:31:18,229 --> 00:31:25,570
boxers and we switch among all the same
time we can compute all possible seats

293
00:31:25,570 --> 00:31:30,658
of the pseudo random generation I do to
us

294
00:31:30,659 --> 00:31:42,299
works so little girl goes back to
brute-force actually she broke into

295
00:31:42,299 --> 00:31:52,289
massage my mother so just that we have
run on the dollar and the device which

296
00:31:52,289 --> 00:32:09,039
boots up and unix time and that's just
so let's talk about private trust who

297
00:32:09,039 --> 00:32:12,769
have you guys know how many actually
artists little the little talk with

298
00:32:12,769 --> 00:32:17,440
secure person he said I should ask this
question to ask the question if you are

299
00:32:17,440 --> 00:32:20,839
the browser and the default install
process and all around them

300
00:32:20,839 --> 00:32:32,869
operation system how many seats do you
trust was a boy opinions 100 but I trust

301
00:32:32,869 --> 00:32:42,860
what a couple dozen 100 couple dozens
what if I think

302
00:32:42,860 --> 00:32:48,840
300 alright so we go to hundreds of
first- second- and 300 cattle

303
00:32:48,840 --> 00:32:53,389
well it's not probably not accept 300
but we have a few hundred and probably

304
00:32:53,390 --> 00:33:00,100
to us we trust a few hundred so you
trust a few hundred companies you never

305
00:33:00,100 --> 00:33:07,209
had a business both i mean you trust
like a few hundred dollar and Rudy pool

306
00:33:07,210 --> 00:33:12,510
companies you don't know which country
they are which lost due to have a good

307
00:33:12,510 --> 00:33:17,379
data protection law and you go far
enough for many people go far enough to

308
00:33:17,380 --> 00:33:21,900
say oh well I need a start for my
website so I go to it all around on that

309
00:33:21,900 --> 00:33:29,179
and I create my private key and even
better

310
00:33:29,179 --> 00:33:34,380
going back to implementation plus there
was before they had engine x10 think

311
00:33:34,380 --> 00:33:41,070
what's called a paci remember back in
the days and they like I don't put up

312
00:33:41,070 --> 00:33:48,870
you have a password so if you look up to
in the internet you'll find alternative

313
00:33:48,870 --> 00:33:52,189
forms where people find out how to
remove the password of your search

314
00:33:52,190 --> 00:33:56,570
because otherwise I put someone's .
check it out i mean that's awesome

315
00:33:56,570 --> 00:34:06,250
that's the best Nicole you can get so
why not trust well who's using talk to

316
00:34:06,250 --> 00:34:15,270
you guys who are not well-known network
right cool why do you store just to

317
00:34:15,270 --> 00:34:16,159
check it out

318
00:34:16,159 --> 00:34:24,149
other opinions why do you store

319
00:34:24,149 --> 00:34:30,239
ok so often pawn

320
00:34:30,239 --> 00:34:34,799
or something real pong

321
00:34:34,800 --> 00:34:43,520
what else could use your store for ya
buying drugs but kind since I'm not a

322
00:34:43,520 --> 00:34:52,210
couple bomb I'm not going to tell you it
was something good I'm sorry ok next

323
00:34:52,210 --> 00:34:55,740
question what makes you trust people and
machines you don't know if you don't

324
00:34:55,739 --> 00:35:01,669
have control off I mean since your store
you trust foreign people you never met

325
00:35:01,670 --> 00:35:14,680
and you don't have control of the
machines so what makes a trusting guy

326
00:35:14,680 --> 00:35:27,180
right did you know that about 58 percent
of all the tour notes are vulnerable to

327
00:35:27,180 --> 00:35:31,319
the timing attack which is heavily used
by our friends in Great Britain United

328
00:35:31,320 --> 00:35:40,890
States on everybody else was playing
perfect timing attacks no yes yes that's

329
00:35:40,890 --> 00:35:48,980
a paper which is came out a few days ago
and I figure it's well we have a problem

330
00:35:48,980 --> 00:35:56,430
of timing attacks a serious one and set
up my cat week water softener to our

331
00:35:56,430 --> 00:36:00,759
network is monitor what about the timing
attack what does anybody know how the

332
00:36:00,760 --> 00:36:06,500
timing attack works well

333
00:36:06,500 --> 00:36:13,029
timing attack the attacker just need to
take control of an accident we note

334
00:36:13,029 --> 00:36:21,939
help you can do some statistical
analysis and find where you from the

335
00:36:21,939 --> 00:36:27,519
matter of minutes there's a new there's
a new paper out which cost measuring and

336
00:36:27,519 --> 00:36:29,158
migrating

337
00:36:29,159 --> 00:36:36,239
advertise what advisories I can store
and others former University I will put

338
00:36:36,239 --> 00:36:40,119
us under the notes when I when I upload
just because it's quite fresh the paper

339
00:36:40,119 --> 00:36:50,339
will put a link up and actually created
by some people from United States which

340
00:36:50,339 --> 00:36:56,880
reduces the problem 25.8% but the way to
us mates you will not be safe against

341
00:36:56,880 --> 00:37:03,179
timing attacks which moves even though
it's reduced to 5% of 6% you still

342
00:37:03,179 --> 00:37:07,880
vulnerable again tight timing attacks
that's just what what it is so if you

343
00:37:07,880 --> 00:37:09,909
want to hide your identity

344
00:37:09,909 --> 00:37:14,819
maybe twice not anymore what you want to
use maybe you just want some service in

345
00:37:14,819 --> 00:37:18,569
China maybe maybe you're just a stroll
down for iOS four intolerant of my

346
00:37:18,569 --> 00:37:25,290
country and

347
00:37:25,290 --> 00:38:07,000
later actually incorporated into the
problem always after while we don't know

348
00:38:07,000 --> 00:39:02,400
the whole problem child she's so if you
look at my location it's not in revoking

349
00:39:02,400 --> 00:39:11,780
that was that was so the idea is
actually two first verify the Trust's

350
00:39:11,780 --> 00:39:16,620
before we start exchanging keys we
should probably a first barry five-word

351
00:39:16,620 --> 00:39:25,040
is what ideas why that is maybe the
question and

352
00:39:25,040 --> 00:39:31,870
stake so there are so few people from
Germany they have a little company right

353
00:39:31,870 --> 00:39:34,440
now packing ideas

354
00:39:34,440 --> 00:39:38,810
mainly for my facts they work on
actually on new model where you can

355
00:39:38,810 --> 00:39:47,350
verify before you trust me i'm not i'm
not entitled to speak about it yet

356
00:39:47,350 --> 00:39:52,150
because they cannot deliver but there
will be some in the pipe we all should

357
00:39:52,150 --> 00:39:58,430
ask yourself about the trust we all
should ask yourself do you really trust

358
00:39:58,430 --> 00:40:02,890
us instant we really trust us over even
though the service run by a friend of

359
00:40:02,890 --> 00:40:08,779
you the question is does he maintains
the stuff right does he have a good seat

360
00:40:08,780 --> 00:40:15,660
does he generates two keys properly all
these questions on stake of it starts

361
00:40:15,660 --> 00:40:22,799
with trust so I think this very moment
are crypto leaks a lot on implementation

362
00:40:22,800 --> 00:40:29,160
floss be computers with analyzing and
the trust model we should definitely

363
00:40:29,160 --> 00:40:33,069
working out west model to make things
better

364
00:40:33,070 --> 00:40:43,960
have to say trust me i'm open for
questions

365
00:40:43,960 --> 00:41:01,089
no questions there yes if you have a
trustworthy notes that would help the

366
00:41:01,089 --> 00:41:01,599
problem

367
00:41:01,599 --> 00:41:07,980
timing attack I think it's like 58
percent of the notes belong to the rest

368
00:41:07,980 --> 00:41:20,510
of the Chinese government so it's just a
question you trust more yeah that's

369
00:41:20,510 --> 00:41:26,670
right that's actually what a new client
is doing I guess but you need to read

370
00:41:26,670 --> 00:41:31,690
the paper first I think ago answer your
question I was put up the paperwork also

371
00:41:31,690 --> 00:41:36,859
some my research I should be more
careful to talk about it

372
00:41:36,859 --> 00:41:41,819
about research other people that so I
think if you read the paper will answer

373
00:41:41,820 --> 00:41:51,780
your question I will put up the link on
the question how many serious do you

374
00:41:51,780 --> 00:41:53,320
trust in your browser

375
00:41:53,320 --> 00:42:03,030
me actually revoked a lot of kicked out
a lot of serious because you cannot

376
00:42:03,030 --> 00:42:07,550
really trust them it's a but it's all
about always a hassle with links to to

377
00:42:07,550 --> 00:42:10,260
do it anyway

378
00:42:10,260 --> 00:42:18,730
GPS is always a hassle understand your
last name is an evil bomb but still

379
00:42:18,730 --> 00:42:24,660
wouldn't artificial delays help fight
that I'm attacked think that's what I

380
00:42:24,660 --> 00:42:29,920
try to do I'm not totally sure what it
is right now cause I just read the paper

381
00:42:29,920 --> 00:42:32,510
yesterday

382
00:42:32,510 --> 00:42:37,430
but before I before I make comments
about this paper I prefer to read a

383
00:42:37,430 --> 00:42:42,740
twenty times and fully understand what
the personals doing good commander

384
00:42:42,740 --> 00:42:51,220
otherwise our first have to do with deep
research before talk about it do you

385
00:42:51,220 --> 00:42:58,859
think they're really coming down the
model of trust in the internet because

386
00:42:58,860 --> 00:43:06,530
everything in life you have to trust you
trust that when he woke the pavement you

387
00:43:06,530 --> 00:43:13,660
want to get to ride bike bike are you
trust Dr the hospital and so on and so

388
00:43:13,660 --> 00:43:20,299
on so on so life entirely realize
somehow contrast to the people who don't

389
00:43:20,300 --> 00:43:26,270
know discussing the internet and was
certificates is just one presentation of

390
00:43:26,270 --> 00:43:34,120
trust in real life you choose to talk to
you go to that doesn't mean I mean let's

391
00:43:34,120 --> 00:43:39,569
say you're going to a dentist he has a
very very good reputation which you take

392
00:43:39,570 --> 00:43:46,260
as a model of trust for yourself and
there's another example doctor who has a

393
00:43:46,260 --> 00:43:51,400
really bad reputation because it always
hurts right you would choose your model

394
00:43:51,400 --> 00:43:54,920
of trust and go to the good doctor but
the good doctor at dr. they play

395
00:43:54,920 --> 00:43:58,710
together goals so we could talk to a
trusted that doctor about dr. trust the

396
00:43:58,710 --> 00:44:03,700
good doctor by implication that means
just because there's a trust in between

397
00:44:03,700 --> 00:44:07,609
those guys and it's an implemented into
your chief you can go to both of them

398
00:44:07,610 --> 00:44:16,920
what I'm saying it's not not trusting
I'm saying very fiber for trust and

399
00:44:16,920 --> 00:44:22,650
trust model in the moment was like this
you trust all of these opportunities

400
00:44:22,650 --> 00:44:28,450
need to trust each other otherwise I
wouldn't work and look at Apple's they

401
00:44:28,450 --> 00:44:33,000
like to keynote long did it take to take
it out i three weeks or two weeks until

402
00:44:33,000 --> 00:44:38,640
he took it out of their trust chain I
mean the shit out of it all look at look

403
00:44:38,640 --> 00:44:44,500
at Apple you know this is little devices
you might hear them a call I think and

404
00:44:44,500 --> 00:44:51,730
they establish a connection to whatever
server in the cloud and to some ass

405
00:44:51,730 --> 00:44:55,150
contralateral that the devices control
you know who don't have control over

406
00:44:55,150 --> 00:45:01,860
device diluted with owner who paid for
it that's not a model of trust me we

407
00:45:01,860 --> 00:45:09,820
should use we should verify or sources
and maybe reduce or sources because they

408
00:45:09,820 --> 00:45:16,470
are countries they have something like a
Patriot Act III some countries to do

409
00:45:16,470 --> 00:45:23,779
something like the US and by this I
company may be entitled or even asked to

410
00:45:23,780 --> 00:45:29,080
publish private information because they
are some people became guns and how to

411
00:45:29,080 --> 00:45:38,610
use them and and now we post about it so
if you trust goes to a to a comment or

412
00:45:38,610 --> 00:45:42,500
two to a company who works in the
government is pressed by the law to give

413
00:45:42,500 --> 00:45:49,590
out information is your trust company
yes no no

414
00:45:49,590 --> 00:45:56,060
for the government yes but maybe not for
you because I unclipped at all it

415
00:45:56,060 --> 00:46:12,000
doesn't mean it is not but anyway it has
your life why do we want our equipment

416
00:46:12,000 --> 00:46:17,730
why do you want to encrypt there's
another case so if you want to use your

417
00:46:17,730 --> 00:46:26,980
information to go out why not posting on
twitter piston it's much easier but not

418
00:46:26,980 --> 00:46:34,490
but maybe you want to the final but
other two but you see where I'm going to

419
00:46:34,490 --> 00:46:44,220
you need to verify who you trust more
questions yes you have or maybe you know

420
00:46:44,220 --> 00:46:50,540
where we can find some advice on which
says should not be trusted at all like

421
00:46:50,540 --> 00:46:54,440
those that are cheney's so that we don't
have to do it by hand

422
00:46:54,440 --> 00:47:00,770
going through hundreds of CA's an hour
I'll post some recommendations in the

423
00:47:00,770 --> 00:47:06,280
holes and the notes but also gonna be
recommendations because of who I am that

424
00:47:06,280 --> 00:47:13,650
you trust me what I recommend you should
actually stick with these guys for

425
00:47:13,650 --> 00:47:18,780
package and you should follow up report
a doing a published maybe next year or

426
00:47:18,780 --> 00:47:26,250
so they say on it since 2013 I think he
will do a lot of talks about it and you

427
00:47:26,250 --> 00:47:32,370
should stick with those guys on the
community likes a community wanna make

428
00:47:32,370 --> 00:47:37,720
programs parenting 101 to the packy does
wanna make security better so maybe we

429
00:47:37,720 --> 00:47:42,370
should just stick with those guys and
find out what you gonna do question was

430
00:47:42,370 --> 00:47:49,200
a scalable and he said yes so maybe we
should check this out I gotta post a

431
00:47:49,200 --> 00:47:54,640
list where smart people that you cannot
trustees on thursday saw that and that's

432
00:47:54,640 --> 00:47:55,210
reason

433
00:47:55,210 --> 00:48:01,560
lots you're free to implement or don't
implemented just how you feel more

434
00:48:01,560 --> 00:48:09,660
questions so that ladies and gentlemen
that's a great audience I think you're

435
00:48:09,660 --> 00:48:12,359
firing me wish you a nice lunch