0 00:00:00,000 --> 00:00:30,000 Dear viewer, these subtitles were generated by a machine via the service Trint and therefore are (very) buggy. If you are capable, please help us to create good quality subtitles: https://c3subtitles.de/talk/597 Thanks! 1 00:00:09,270 --> 00:00:11,519 And now are we going to have a discussion 2 00:00:11,520 --> 00:00:13,769 about the crypto wars, which is this 3 00:00:13,770 --> 00:00:15,749 ongoing conflict, about the fact that 4 00:00:15,750 --> 00:00:16,979 some governments seem to be very 5 00:00:16,980 --> 00:00:19,259 insistent on not allowing people 6 00:00:19,260 --> 00:00:21,029 to encrypt the data? 7 00:00:21,030 --> 00:00:22,529 But luckily, there's some people fighting 8 00:00:22,530 --> 00:00:23,909 against them. And among them there's the 9 00:00:23,910 --> 00:00:25,949 electronic frontier, the Electronic 10 00:00:25,950 --> 00:00:28,589 Frontier Foundation, the EFF. 11 00:00:28,590 --> 00:00:30,500 Yeah, thank you. 12 00:00:35,100 --> 00:00:36,539 Which apparently also dabbles in 13 00:00:36,540 --> 00:00:37,859 publishing, I don't know if you noticed 14 00:00:37,860 --> 00:00:39,929 as they released a book, a 15 00:00:39,930 --> 00:00:41,999 science fiction collection recently. 16 00:00:42,000 --> 00:00:44,159 It's really strange, but it's a very good 17 00:00:44,160 --> 00:00:45,389 collection. You should read it. 18 00:00:45,390 --> 00:00:47,529 And now, please welcome 19 00:00:47,530 --> 00:00:49,390 Scott Opsahl from the EFSF. 20 00:00:51,080 --> 00:00:53,309 Oh, thank you. 21 00:00:53,310 --> 00:00:54,749 Thank you very much. 22 00:00:54,750 --> 00:00:57,089 Is my name is Kurt Opsahl, 23 00:00:57,090 --> 00:00:58,799 I'm the deputy executive director and 24 00:00:58,800 --> 00:01:00,419 general counsel, Electronic Frontier 25 00:01:00,420 --> 00:01:01,829 Foundation. 26 00:01:01,830 --> 00:01:03,539 It's great pleasure to be back here at 27 00:01:03,540 --> 00:01:05,789 C.C.C.. And thank you all for 28 00:01:05,790 --> 00:01:07,229 for coming. I know it's a little early in 29 00:01:07,230 --> 00:01:08,969 the day, so thank you for making the 30 00:01:08,970 --> 00:01:10,559 effort to come here today. 31 00:01:10,560 --> 00:01:12,449 We're going to talk about the crypto 32 00:01:12,450 --> 00:01:15,419 wars, what we call part two, 33 00:01:15,420 --> 00:01:17,759 the crypto wars. Part one was 34 00:01:17,760 --> 00:01:20,039 about 20 years ago 35 00:01:20,040 --> 00:01:22,379 and now it's back. 36 00:01:22,380 --> 00:01:24,749 And we fought pretty hard back then 37 00:01:24,750 --> 00:01:26,879 against attempts to limit, suppress and 38 00:01:26,880 --> 00:01:28,709 cripple encryption. 39 00:01:28,710 --> 00:01:30,539 My organization, the Electronic Frontier 40 00:01:30,540 --> 00:01:32,849 Foundation, was on the front lines 41 00:01:32,850 --> 00:01:35,189 fighting throughout the 90s to 42 00:01:35,190 --> 00:01:37,319 help preserve the ability for people to 43 00:01:37,320 --> 00:01:39,629 use, develop and publish strong 44 00:01:39,630 --> 00:01:40,629 encryption. 45 00:01:40,630 --> 00:01:43,859 It was a long struggle, took many years. 46 00:01:43,860 --> 00:01:46,559 But finally, by by the end of the 90s, 47 00:01:46,560 --> 00:01:48,209 encryption prevailed. 48 00:01:48,210 --> 00:01:50,519 And we thought, at least for a time, 49 00:01:50,520 --> 00:01:52,379 we would have the freedom to continue to 50 00:01:52,380 --> 00:01:53,969 use the strongest encryption as we 51 00:01:53,970 --> 00:01:55,379 possibly could. 52 00:01:55,380 --> 00:01:57,449 But now the governments are at it 53 00:01:57,450 --> 00:01:58,379 again. 54 00:01:58,380 --> 00:02:00,449 They come back to try and weaken 55 00:02:00,450 --> 00:02:02,760 or destroy crypto, make it harder to use. 56 00:02:03,990 --> 00:02:05,669 So today we're going to talk about a 57 00:02:05,670 --> 00:02:06,839 little bit of the history 58 00:02:08,039 --> 00:02:10,168 going back to the original crypto 59 00:02:10,169 --> 00:02:12,509 wars, a little bit about what governments 60 00:02:12,510 --> 00:02:14,609 are now doing to try and reignite 61 00:02:14,610 --> 00:02:15,749 those wars. 62 00:02:15,750 --> 00:02:17,879 And then some of the the arguments 63 00:02:17,880 --> 00:02:20,579 that you can make to help fight back 64 00:02:20,580 --> 00:02:22,709 and preserve your access to strong 65 00:02:22,710 --> 00:02:23,710 encryption. 66 00:02:24,960 --> 00:02:25,859 So we'll start with a bit of the 67 00:02:25,860 --> 00:02:27,599 background. 68 00:02:27,600 --> 00:02:29,669 You know, in the earlier times, 69 00:02:29,670 --> 00:02:31,169 you know, through the Cold War, even 70 00:02:31,170 --> 00:02:33,569 before encryption was generally 71 00:02:33,570 --> 00:02:35,669 considered to be a military technology. 72 00:02:35,670 --> 00:02:37,889 Not always. I mean, there was actually 73 00:02:37,890 --> 00:02:39,809 a commercial version of the Enigma 74 00:02:39,810 --> 00:02:42,569 machine in the 1920s 75 00:02:42,570 --> 00:02:44,159 and 30s. 76 00:02:44,160 --> 00:02:46,259 But by and large, it was 77 00:02:46,260 --> 00:02:48,149 used most commonly in the military and 78 00:02:48,150 --> 00:02:50,549 thought of as a military technology 79 00:02:50,550 --> 00:02:52,799 by about the 70s 80 00:02:52,800 --> 00:02:54,599 that came out with the standard of the 81 00:02:54,600 --> 00:02:57,509 data encryption standard disks. 82 00:02:57,510 --> 00:03:00,329 And this started to popularize 83 00:03:00,330 --> 00:03:01,739 encryption to start to be used more 84 00:03:01,740 --> 00:03:03,209 frequently. 85 00:03:03,210 --> 00:03:05,399 It also sort of help 86 00:03:05,400 --> 00:03:07,109 jumpstart some of the fields of crypto 87 00:03:07,110 --> 00:03:09,359 analysis. There was an algorithm that 88 00:03:09,360 --> 00:03:11,159 people could start to look at and 89 00:03:11,160 --> 00:03:14,189 intensely try to see if there were flaws. 90 00:03:14,190 --> 00:03:15,479 Turned out there were some flaws, but we 91 00:03:15,480 --> 00:03:18,689 didn't find that out till till later. 92 00:03:18,690 --> 00:03:21,869 In the late 70s, RSA 93 00:03:21,870 --> 00:03:24,269 found a way to implement the Diffie 94 00:03:24,270 --> 00:03:26,729 Hellman public key exchange. 95 00:03:26,730 --> 00:03:29,069 This was a great advance in technology 96 00:03:29,070 --> 00:03:30,599 allowing people to send encrypted 97 00:03:30,600 --> 00:03:32,819 communications when they weren't 98 00:03:32,820 --> 00:03:33,820 able to 99 00:03:35,160 --> 00:03:37,979 have a side channel to exchange the key. 100 00:03:37,980 --> 00:03:38,939 This was very important in the 101 00:03:38,940 --> 00:03:41,669 development of cryptography. 102 00:03:41,670 --> 00:03:43,939 By the early 90s, we had 103 00:03:43,940 --> 00:03:45,719 the pretty good privacy and email 104 00:03:45,720 --> 00:03:47,729 encryption program that most of you are 105 00:03:47,730 --> 00:03:49,799 probably familiar with, allowing 106 00:03:49,800 --> 00:03:51,509 people to have ended encrypted email 107 00:03:51,510 --> 00:03:53,489 communications. 108 00:03:53,490 --> 00:03:55,889 And by the mid 90s we had 109 00:03:55,890 --> 00:03:58,019 secured socket layer introduced by 110 00:03:58,020 --> 00:04:00,149 the Netscape for its Navigator 111 00:04:00,150 --> 00:04:02,339 browser program, allowing servers 112 00:04:02,340 --> 00:04:04,349 to communicate with encryption and 113 00:04:04,350 --> 00:04:06,629 authentication between the server 114 00:04:06,630 --> 00:04:07,630 and the browser. 115 00:04:09,010 --> 00:04:11,139 But this presented some some challenges, 116 00:04:11,140 --> 00:04:13,419 and one of those challenges was 117 00:04:13,420 --> 00:04:15,099 what is what does the munition? 118 00:04:15,100 --> 00:04:16,838 So I have up on the screen two types of 119 00:04:16,839 --> 00:04:19,148 munitions on the right, a more familiar 120 00:04:19,149 --> 00:04:21,219 one, the tank, which 121 00:04:21,220 --> 00:04:22,419 has export controls. 122 00:04:22,420 --> 00:04:24,759 And many people will find that sensible, 123 00:04:24,760 --> 00:04:26,169 that before you're going to ship a tank 124 00:04:26,170 --> 00:04:28,239 to people, you should, I know, know 125 00:04:28,240 --> 00:04:29,529 what you're doing. 126 00:04:29,530 --> 00:04:31,629 And on the on the left, we have 127 00:04:31,630 --> 00:04:33,639 the Netscape Navigator. 128 00:04:33,640 --> 00:04:36,189 This is the the international version 129 00:04:36,190 --> 00:04:38,249 approved for export. 130 00:04:39,250 --> 00:04:40,510 And at the time, 131 00:04:42,070 --> 00:04:45,039 the standard for encryption was 128 132 00:04:45,040 --> 00:04:47,919 bit secure and socket layer 133 00:04:47,920 --> 00:04:50,049 for the domestic version, while the 134 00:04:50,050 --> 00:04:52,119 international the export version 135 00:04:52,120 --> 00:04:54,369 was limited to 40 136 00:04:54,370 --> 00:04:56,529 bit encryption, which 137 00:04:56,530 --> 00:04:58,449 turns out could be could be broken in 138 00:04:58,450 --> 00:05:00,669 days. And the idea behind this limitation 139 00:05:00,670 --> 00:05:02,739 was in part to make sure that 140 00:05:02,740 --> 00:05:05,049 organizations like the the NSA 141 00:05:05,050 --> 00:05:06,879 could easily break the encryption, that 142 00:05:06,880 --> 00:05:08,289 we weren't exporting something that 143 00:05:08,290 --> 00:05:10,239 couldn't be broken by our security 144 00:05:10,240 --> 00:05:11,229 services. 145 00:05:11,230 --> 00:05:13,059 But this this created some sort of, you 146 00:05:13,060 --> 00:05:15,339 know, relative craziness. 147 00:05:15,340 --> 00:05:17,439 That encryption 148 00:05:17,440 --> 00:05:19,659 that was developed overseas 149 00:05:19,660 --> 00:05:21,219 could be stronger. 150 00:05:21,220 --> 00:05:23,409 Cryptography conferences are to be held 151 00:05:23,410 --> 00:05:25,479 in places like Caribbean islands 152 00:05:25,480 --> 00:05:27,669 to get around these encryption. 153 00:05:27,670 --> 00:05:29,709 And there were some legal challenges to 154 00:05:29,710 --> 00:05:32,180 these export regulations. 155 00:05:33,370 --> 00:05:35,529 Now, one of the legacies that 156 00:05:35,530 --> 00:05:37,629 came out of these export 157 00:05:37,630 --> 00:05:39,970 grade encryption we just saw actually 158 00:05:41,020 --> 00:05:43,659 this year illustrated 159 00:05:43,660 --> 00:05:46,479 through the freak and logjam attacks. 160 00:05:46,480 --> 00:05:47,499 I'm going to talk about this very 161 00:05:47,500 --> 00:05:48,699 briefly. There was actually a great 162 00:05:48,700 --> 00:05:50,829 session yesterday from 163 00:05:50,830 --> 00:05:53,049 Alex Halderman and not a Hettinger to go 164 00:05:53,050 --> 00:05:54,189 through them in great detail. 165 00:05:54,190 --> 00:05:55,989 I suggest you watch that if you haven't 166 00:05:55,990 --> 00:05:57,309 already. 167 00:05:57,310 --> 00:05:59,649 But the the takeaway from this for 168 00:06:00,820 --> 00:06:03,219 for the crypto wars was that the 169 00:06:03,220 --> 00:06:06,219 legacy of export grade encryption 170 00:06:06,220 --> 00:06:08,349 persisted 20 years later 171 00:06:08,350 --> 00:06:11,109 so that people were able to convince 172 00:06:11,110 --> 00:06:13,359 browsers service to downgrade 173 00:06:13,360 --> 00:06:15,609 to the export grade keys, which 174 00:06:15,610 --> 00:06:17,799 were designed to allow the 175 00:06:17,800 --> 00:06:20,889 NSA to be able to break them. 176 00:06:20,890 --> 00:06:22,479 But it didn't really have the foresight 177 00:06:22,480 --> 00:06:24,099 to think about how much computing power 178 00:06:24,100 --> 00:06:25,539 would change over the years. 179 00:06:25,540 --> 00:06:27,609 So perhaps in the 180 00:06:27,610 --> 00:06:29,409 1990s, when they were trying to come up 181 00:06:29,410 --> 00:06:31,599 with these export grade levels, 182 00:06:31,600 --> 00:06:33,159 they thought that, well, only only the 183 00:06:33,160 --> 00:06:35,319 NSA would have the computing power to do 184 00:06:35,320 --> 00:06:37,479 this. And so this was a wise 185 00:06:37,480 --> 00:06:39,159 idea, but it turned out not to be a very 186 00:06:39,160 --> 00:06:40,269 wise idea. 187 00:06:40,270 --> 00:06:42,549 Twenty years later, when you can go on 188 00:06:42,550 --> 00:06:45,159 to Amazon and for 189 00:06:45,160 --> 00:06:47,229 75 dollars and a couple of hours of 190 00:06:47,230 --> 00:06:49,659 time, be able to break some of these 191 00:06:49,660 --> 00:06:51,369 export grade encryptions. 192 00:06:51,370 --> 00:06:53,349 So we're still having sort of casualties 193 00:06:53,350 --> 00:06:55,629 from the first crypto 194 00:06:55,630 --> 00:06:56,630 wars today. 195 00:06:58,600 --> 00:07:00,219 During the 90s there, we had a 196 00:07:00,220 --> 00:07:02,529 challenger, Dan Bernstein, 197 00:07:02,530 --> 00:07:05,229 challenge export control of his snuffle 198 00:07:05,230 --> 00:07:06,729 crypto program. 199 00:07:06,730 --> 00:07:09,729 This was a case that the FBI helped and 200 00:07:09,730 --> 00:07:11,889 took it to the courts and 201 00:07:11,890 --> 00:07:13,509 had some success. 202 00:07:13,510 --> 00:07:15,999 The courts determined that code 203 00:07:16,000 --> 00:07:17,709 was speech, that this was a free 204 00:07:17,710 --> 00:07:19,839 expression, protected act that 205 00:07:19,840 --> 00:07:21,039 you could publish. 206 00:07:21,040 --> 00:07:22,719 You're the source code. 207 00:07:22,720 --> 00:07:24,519 You could publish your algorithms. 208 00:07:24,520 --> 00:07:27,099 And there shouldn't be treated 209 00:07:27,100 --> 00:07:29,199 as a munition, but rather 210 00:07:29,200 --> 00:07:30,549 thought of as speech. 211 00:07:30,550 --> 00:07:32,629 And out of the the appeal from that case, 212 00:07:32,630 --> 00:07:34,569 we had a great quote that that came from 213 00:07:34,570 --> 00:07:35,829 the court showing that the courts were 214 00:07:35,830 --> 00:07:37,929 recognizing both the freedom of 215 00:07:37,930 --> 00:07:41,109 expression rights here, but also 216 00:07:41,110 --> 00:07:43,299 the privacy value that 217 00:07:43,300 --> 00:07:45,159 they recognized that the availability of 218 00:07:45,160 --> 00:07:47,139 strong encryption would help people 219 00:07:47,140 --> 00:07:49,359 protect their privacy, implicating not 220 00:07:49,360 --> 00:07:51,429 only First Amendment rights, 221 00:07:51,430 --> 00:07:53,229 but their thinking about the right to 222 00:07:53,230 --> 00:07:55,689 receive information as recipients of 223 00:07:55,690 --> 00:07:57,219 encryptions bounty. 224 00:07:57,220 --> 00:07:59,319 So this was a very strong blow 225 00:07:59,320 --> 00:08:01,419 that helped enable 226 00:08:01,420 --> 00:08:03,309 encryption moving forward. 227 00:08:03,310 --> 00:08:04,869 Another legacy of the 90s. 228 00:08:04,870 --> 00:08:06,879 It was the Clipper chip. 229 00:08:06,880 --> 00:08:09,099 This was a chip set. 230 00:08:09,100 --> 00:08:11,679 It was designed for voice communications 231 00:08:11,680 --> 00:08:13,779 to encrypt voice communications 232 00:08:13,780 --> 00:08:15,999 with skipjack encryption 233 00:08:16,000 --> 00:08:16,989 algorithm. 234 00:08:16,990 --> 00:08:19,449 And then it included a back door 235 00:08:19,450 --> 00:08:21,549 with a key that was 236 00:08:21,550 --> 00:08:23,649 supposed to be placed in escrow 237 00:08:23,650 --> 00:08:25,149 so that when the government came 238 00:08:25,150 --> 00:08:27,099 knocking, they could get the escrowed 239 00:08:27,100 --> 00:08:29,469 key, decrypt the communications 240 00:08:29,470 --> 00:08:31,239 and find out what people said. 241 00:08:31,240 --> 00:08:32,798 And they wanted to have people install 242 00:08:32,799 --> 00:08:35,529 this chip on all their phones. 243 00:08:35,530 --> 00:08:37,629 And we and many others 244 00:08:37,630 --> 00:08:39,788 pointed out to them back doors can be 245 00:08:39,789 --> 00:08:41,019 very dangerous. 246 00:08:41,020 --> 00:08:43,089 Even a small flaw in 247 00:08:43,090 --> 00:08:45,159 a crypto system can ultimately 248 00:08:45,160 --> 00:08:47,259 end up being disastrous. 249 00:08:47,260 --> 00:08:49,569 And as it turned out, the Clippers 250 00:08:49,570 --> 00:08:51,549 did have some flaws. 251 00:08:51,550 --> 00:08:53,769 By the 1994, Matt Blaze 252 00:08:53,770 --> 00:08:56,349 was able to determine 253 00:08:56,350 --> 00:08:58,599 that the law enforcement access 254 00:08:58,600 --> 00:09:00,789 field contained in the information 255 00:09:00,790 --> 00:09:02,539 needed to recover the key. 256 00:09:03,550 --> 00:09:05,619 So this made the key escrow, 257 00:09:05,620 --> 00:09:07,869 not just a back door for the NSA, 258 00:09:07,870 --> 00:09:09,999 but a back door to for other people to 259 00:09:10,000 --> 00:09:11,080 access, as well 260 00:09:12,310 --> 00:09:14,379 as later an attack that was able to 261 00:09:14,380 --> 00:09:17,199 bypass the escrow. 262 00:09:17,200 --> 00:09:19,659 So Kliper ended up being widely condemned 263 00:09:19,660 --> 00:09:22,539 and eventually was 264 00:09:22,540 --> 00:09:24,699 sort of universally considered to be a 265 00:09:24,700 --> 00:09:25,700 bad idea. 266 00:09:28,010 --> 00:09:30,259 So when we look back over 267 00:09:30,260 --> 00:09:32,479 the 1990s, policy debate 268 00:09:32,480 --> 00:09:34,609 is actually very eerily similar 269 00:09:34,610 --> 00:09:36,319 to today, a lot of the same talking 270 00:09:36,320 --> 00:09:38,509 points are being used, a lot 271 00:09:38,510 --> 00:09:39,740 of the same arguments, 272 00:09:41,210 --> 00:09:42,679 for example. And they're actually many 273 00:09:42,680 --> 00:09:44,480 examples. I'm just going to pick a few. 274 00:09:45,530 --> 00:09:47,689 But in 1997, 275 00:09:47,690 --> 00:09:49,099 the director of the Federal Bureau of 276 00:09:49,100 --> 00:09:51,199 Investigation said 277 00:09:51,200 --> 00:09:52,969 strong encryption. It's great. 278 00:09:52,970 --> 00:09:54,379 We love it. 279 00:09:54,380 --> 00:09:56,389 We just want to make sure that there's a 280 00:09:56,390 --> 00:09:58,159 key that's available so that we can get 281 00:09:58,160 --> 00:09:59,160 access to it. 282 00:10:00,140 --> 00:10:02,749 Fast forward to the 2000s 283 00:10:02,750 --> 00:10:04,849 and the FBI's general counsel 284 00:10:04,850 --> 00:10:07,459 says, yeah, great strong encryption, 285 00:10:07,460 --> 00:10:08,899 but they just need to be able to find a 286 00:10:08,900 --> 00:10:11,239 way to give us the plain text. 287 00:10:11,240 --> 00:10:13,309 So we're considering the the sort of 288 00:10:13,310 --> 00:10:15,589 same rhetoric that 289 00:10:15,590 --> 00:10:17,719 they realized that perhaps saying we 290 00:10:17,720 --> 00:10:20,059 want weak encryption doesn't sound right. 291 00:10:20,060 --> 00:10:22,849 People are people not receptive to that. 292 00:10:22,850 --> 00:10:24,589 So they say great, strong encryption. 293 00:10:24,590 --> 00:10:25,669 We're all for it. 294 00:10:25,670 --> 00:10:27,320 It just has to have a giant hole in it. 295 00:10:30,440 --> 00:10:32,629 So what happened in the in 296 00:10:32,630 --> 00:10:35,059 the 2010s that brought this 297 00:10:35,060 --> 00:10:36,739 the second round of the crypto wars? 298 00:10:36,740 --> 00:10:38,089 Why the why did the government sort of 299 00:10:38,090 --> 00:10:40,549 start attacking encryption again? 300 00:10:40,550 --> 00:10:42,619 The major shift is the 301 00:10:42,620 --> 00:10:45,229 development of widespread encryption. 302 00:10:46,820 --> 00:10:48,889 The key event that really started off 303 00:10:48,890 --> 00:10:51,199 government rhetoric came around the time 304 00:10:51,200 --> 00:10:53,299 that that first Apple and shortly 305 00:10:53,300 --> 00:10:54,949 thereafter, Google said that they were 306 00:10:54,950 --> 00:10:57,379 going to have encryption on 307 00:10:57,380 --> 00:10:59,239 their phones for the for the data stored 308 00:10:59,240 --> 00:11:02,179 on the phone that they couldn't 309 00:11:02,180 --> 00:11:04,069 unlock at the government's request. 310 00:11:04,070 --> 00:11:06,499 So, for example, with iOS seven 311 00:11:06,500 --> 00:11:08,629 and before it had encryption on 312 00:11:08,630 --> 00:11:10,369 there. But if you sent the phone in to 313 00:11:10,370 --> 00:11:12,709 Apple and, you know, brought a warrant 314 00:11:12,710 --> 00:11:15,469 and such, that Apple could 315 00:11:15,470 --> 00:11:17,299 unlock the phone, get access to the 316 00:11:17,300 --> 00:11:18,349 information. 317 00:11:18,350 --> 00:11:20,599 And they said, well, with the new iOS 318 00:11:20,600 --> 00:11:22,639 eight and later they wouldn't be able to 319 00:11:22,640 --> 00:11:24,889 do that, that the only the user 320 00:11:24,890 --> 00:11:26,539 would be able to unlock their their 321 00:11:26,540 --> 00:11:27,859 phones. 322 00:11:27,860 --> 00:11:29,569 At the same time, more and more 323 00:11:29,570 --> 00:11:31,879 commonplace messaging apps started 324 00:11:31,880 --> 00:11:33,379 to have encryption. 325 00:11:33,380 --> 00:11:35,029 I message was one of the early things 326 00:11:35,030 --> 00:11:36,829 actually since the beginning of my 327 00:11:36,830 --> 00:11:39,019 message. It had opportunistic 328 00:11:39,020 --> 00:11:40,309 encryption. So if it was 329 00:11:41,360 --> 00:11:43,579 a text sent from an iPhone to an iPhone, 330 00:11:43,580 --> 00:11:45,769 it would use a message and encrypt 331 00:11:45,770 --> 00:11:47,039 the message. 332 00:11:47,040 --> 00:11:49,219 There also became the text secure 333 00:11:49,220 --> 00:11:51,469 became popularized for Android. 334 00:11:51,470 --> 00:11:53,779 WhatsApp, one of the most popular 335 00:11:53,780 --> 00:11:56,009 communication tools out there with a 336 00:11:56,010 --> 00:11:57,889 at least a billion users. 337 00:11:57,890 --> 00:12:00,109 It started out being unencrypted, 338 00:12:00,110 --> 00:12:02,629 but has started to work with 339 00:12:02,630 --> 00:12:05,359 open whisper systems to add encryption 340 00:12:05,360 --> 00:12:07,849 and encrypt communications of billions of 341 00:12:07,850 --> 00:12:09,199 users. 342 00:12:09,200 --> 00:12:10,969 And this is bringing about more 343 00:12:10,970 --> 00:12:12,439 ubiquitous encryption. 344 00:12:12,440 --> 00:12:14,359 So things that had previously been 345 00:12:14,360 --> 00:12:17,179 limited to more technically sophisticated 346 00:12:17,180 --> 00:12:19,429 users are relatively rare, 347 00:12:19,430 --> 00:12:20,749 something that when you know, when you 348 00:12:20,750 --> 00:12:22,189 look at some of the Snowden documents, 349 00:12:22,190 --> 00:12:24,289 they talk about encryption being used 350 00:12:24,290 --> 00:12:25,789 as a flag to say, well, this is 351 00:12:25,790 --> 00:12:28,129 interesting because it's encrypted 352 00:12:28,130 --> 00:12:29,779 or they say, you know, we will store 353 00:12:29,780 --> 00:12:31,969 encrypted communications until such time 354 00:12:31,970 --> 00:12:34,279 as we're able to decrypt them. 355 00:12:34,280 --> 00:12:35,989 But as more and more communications will 356 00:12:35,990 --> 00:12:37,879 be encrypted, where it was becoming less 357 00:12:37,880 --> 00:12:40,129 unusual, less rare, 358 00:12:40,130 --> 00:12:41,629 this started to become something very 359 00:12:41,630 --> 00:12:44,629 worrisome for the government. 360 00:12:44,630 --> 00:12:47,119 So it started out as 361 00:12:47,120 --> 00:12:49,489 as I said, from the announcements 362 00:12:49,490 --> 00:12:51,799 of having strong encryption 363 00:12:51,800 --> 00:12:53,449 without the ability for companies to 364 00:12:53,450 --> 00:12:55,759 unlock for mobile phones. 365 00:12:55,760 --> 00:12:58,009 Though the conversation quickly moved to 366 00:12:58,010 --> 00:13:00,109 talking about end to end encryption 367 00:13:00,110 --> 00:13:02,659 as a problem, the 368 00:13:02,660 --> 00:13:04,249 prime minister of the United Kingdom, 369 00:13:04,250 --> 00:13:06,769 Cameron, asked rhetorically, 370 00:13:06,770 --> 00:13:08,719 are we going to allow a means of 371 00:13:08,720 --> 00:13:10,639 communication which simply isn't possible 372 00:13:10,640 --> 00:13:12,529 to read? And his answer, of course, was 373 00:13:12,530 --> 00:13:14,419 was no, not if he has his way. 374 00:13:14,420 --> 00:13:16,549 He doesn't want there to be a private 375 00:13:16,550 --> 00:13:18,619 space where you can communicate without 376 00:13:18,620 --> 00:13:20,329 having the government able to look over 377 00:13:20,330 --> 00:13:22,099 your shoulder that there should be no 378 00:13:22,100 --> 00:13:23,100 place to hide. 379 00:13:24,350 --> 00:13:26,839 So they started to strike back 380 00:13:26,840 --> 00:13:28,489 and using some of the tools that we 381 00:13:28,490 --> 00:13:31,639 remember from the first crypto wars, 382 00:13:31,640 --> 00:13:33,619 public and private pressure on the 383 00:13:33,620 --> 00:13:35,089 companies who are supplying these 384 00:13:35,090 --> 00:13:37,279 technologies, public rhetoric 385 00:13:37,280 --> 00:13:39,199 to try and demonize encryption, make it 386 00:13:39,200 --> 00:13:41,719 look bad for the public, 387 00:13:41,720 --> 00:13:43,909 and then by changing that atmosphere, 388 00:13:43,910 --> 00:13:46,399 try and create a more ripe situation for 389 00:13:46,400 --> 00:13:48,229 legislation. 390 00:13:48,230 --> 00:13:50,869 And of course, at the same time, the 391 00:13:50,870 --> 00:13:53,719 the NSA, the JQ 392 00:13:53,720 --> 00:13:55,849 intelligence agencies are working 393 00:13:55,850 --> 00:13:57,949 to use technical attacks 394 00:13:57,950 --> 00:14:00,440 to try to weaken or defeat encryption. 395 00:14:03,310 --> 00:14:05,319 So a lot of the public pressure is on 396 00:14:05,320 --> 00:14:08,079 there. I mean, you know, the FBI director 397 00:14:08,080 --> 00:14:09,849 asks, you know, why would you market 398 00:14:09,850 --> 00:14:11,679 something to allow people to place 399 00:14:11,680 --> 00:14:12,969 themselves beyond the law? 400 00:14:12,970 --> 00:14:14,469 And this is sort of a way of twisting the 401 00:14:14,470 --> 00:14:16,749 conversation, you know, not asking, 402 00:14:16,750 --> 00:14:18,549 you know, why would someone want to have 403 00:14:18,550 --> 00:14:20,529 something that would protect people from 404 00:14:20,530 --> 00:14:22,209 oppressive regimes or would it give them 405 00:14:22,210 --> 00:14:24,889 a private space or allow for 406 00:14:24,890 --> 00:14:26,589 a secure e-commerce? 407 00:14:26,590 --> 00:14:28,389 But sort of discussing this as something 408 00:14:28,390 --> 00:14:30,879 that is only purpose is 409 00:14:30,880 --> 00:14:33,129 to enable criminal activities. 410 00:14:33,130 --> 00:14:35,049 And Cameron to putting the pressure on 411 00:14:35,050 --> 00:14:36,219 saying that companies have a 412 00:14:36,220 --> 00:14:38,379 responsibility to fight terrorism. 413 00:14:38,380 --> 00:14:40,089 And if they you know, if they care if 414 00:14:40,090 --> 00:14:42,369 they're patriotic, they will do as I say 415 00:14:42,370 --> 00:14:44,289 and put in these back doors, put in these 416 00:14:44,290 --> 00:14:46,179 governments ability. 417 00:14:46,180 --> 00:14:47,979 And they're doing this also hard to focus 418 00:14:47,980 --> 00:14:50,379 on the companies with large user 419 00:14:50,380 --> 00:14:52,569 bases saying before 420 00:14:52,570 --> 00:14:53,979 the thing that I was really scared, the 421 00:14:53,980 --> 00:14:56,229 government is sort of the widespread 422 00:14:56,230 --> 00:14:58,869 by default availability of encryption. 423 00:14:58,870 --> 00:15:00,729 So they're trying to get that switched 424 00:15:00,730 --> 00:15:03,129 off for the large companies, 425 00:15:03,130 --> 00:15:05,319 relegating encryption back to 426 00:15:05,320 --> 00:15:07,419 a relatively rare thing that 427 00:15:07,420 --> 00:15:09,279 can be used for for targeting. 428 00:15:10,420 --> 00:15:12,519 They're also trying to be dismissive 429 00:15:12,520 --> 00:15:14,679 of the companies 430 00:15:14,680 --> 00:15:16,989 saying, well, it's only a business model. 431 00:15:16,990 --> 00:15:19,179 It's it's not a technical feature. 432 00:15:19,180 --> 00:15:21,609 It is a marketing pitch. 433 00:15:21,610 --> 00:15:23,049 Well, in some sense, you know, that that 434 00:15:23,050 --> 00:15:25,239 that I think that's great. 435 00:15:25,240 --> 00:15:26,889 Strong encryption should be a marketing 436 00:15:26,890 --> 00:15:29,139 pitch, should be a reason why you're 437 00:15:29,140 --> 00:15:31,269 providing providing the product 438 00:15:31,270 --> 00:15:32,769 to a willing audience. 439 00:15:32,770 --> 00:15:33,729 But it is more than that. 440 00:15:33,730 --> 00:15:36,069 It's also something of principle. 441 00:15:36,070 --> 00:15:37,419 And they're trying to remove the 442 00:15:37,420 --> 00:15:39,549 principle and try and 443 00:15:39,550 --> 00:15:41,859 dismiss the companies as being only 444 00:15:41,860 --> 00:15:42,909 profit motivated. 445 00:15:45,510 --> 00:15:47,669 So they're what they call some 446 00:15:47,670 --> 00:15:49,949 proposals for secure back doors, 447 00:15:49,950 --> 00:15:51,509 and this seems like a contradiction in 448 00:15:51,510 --> 00:15:52,770 terms or a misnomer, 449 00:15:53,850 --> 00:15:55,169 but nevertheless there but a lot of 450 00:15:55,170 --> 00:15:57,539 proposals out there, the most common 451 00:15:57,540 --> 00:15:58,739 is key escrow. 452 00:15:58,740 --> 00:16:01,289 This notion that you have a message 453 00:16:01,290 --> 00:16:03,779 that is sent with a symmetric key, you 454 00:16:03,780 --> 00:16:06,149 encrypt that key twice, once 455 00:16:06,150 --> 00:16:08,309 to the recipients, and they can decrypted 456 00:16:08,310 --> 00:16:10,379 and read it. The escrow agent gets a 457 00:16:10,380 --> 00:16:12,659 copy that is stored so 458 00:16:12,660 --> 00:16:14,909 that at some point later, if they want 459 00:16:14,910 --> 00:16:16,259 to get it, they can get it from the 460 00:16:16,260 --> 00:16:18,449 escrow. Use that key, 461 00:16:18,450 --> 00:16:19,460 read the message. 462 00:16:21,180 --> 00:16:23,549 But this has a number of problems. 463 00:16:23,550 --> 00:16:25,619 If the escrow agents private key 464 00:16:25,620 --> 00:16:27,389 is compromised, it breaks the whole 465 00:16:27,390 --> 00:16:29,939 system. They will be able to use that key 466 00:16:29,940 --> 00:16:32,669 to go back and get messages 467 00:16:32,670 --> 00:16:34,799 for well, as many people as that key 468 00:16:34,800 --> 00:16:36,599 can can unlock. 469 00:16:36,600 --> 00:16:38,690 So it creates a point of failure. 470 00:16:39,960 --> 00:16:42,029 And then if you're using a single 471 00:16:42,030 --> 00:16:44,459 escrow, this can break forward secrecy, 472 00:16:44,460 --> 00:16:46,139 a property that is very useful for 473 00:16:46,140 --> 00:16:47,369 creating strong encryption. 474 00:16:47,370 --> 00:16:49,499 It's a recommended property that I think 475 00:16:49,500 --> 00:16:51,599 we want to have as widespread 476 00:16:51,600 --> 00:16:52,559 as possible. 477 00:16:52,560 --> 00:16:54,809 But if the whole design is to allow 478 00:16:54,810 --> 00:16:57,209 it so that somebody at any point 479 00:16:57,210 --> 00:16:59,399 can get the escrow key, 480 00:16:59,400 --> 00:17:01,289 go back and read all the messages in the 481 00:17:01,290 --> 00:17:04,108 past, this is sort of defeating the point 482 00:17:04,109 --> 00:17:06,358 of forward 483 00:17:06,359 --> 00:17:07,588 secrecy. 484 00:17:07,589 --> 00:17:08,879 Now, you can mitigate some of these 485 00:17:08,880 --> 00:17:11,039 things. Some proposals have talked about 486 00:17:11,040 --> 00:17:13,348 split keys so 487 00:17:13,349 --> 00:17:14,818 that you might have to get more than one 488 00:17:14,819 --> 00:17:16,949 key in order to get it from 489 00:17:16,950 --> 00:17:18,179 the escrow agent. 490 00:17:18,180 --> 00:17:21,239 And, you know, this this adds some 491 00:17:21,240 --> 00:17:23,309 mitigation so that you have to 492 00:17:23,310 --> 00:17:25,229 compromise a couple of different places 493 00:17:25,230 --> 00:17:27,118 to move forward. But all of these things 494 00:17:27,119 --> 00:17:29,249 add complexity and complexity 495 00:17:29,250 --> 00:17:31,409 is the enemy of security. 496 00:17:31,410 --> 00:17:33,179 For every complexity you're adding, 497 00:17:33,180 --> 00:17:35,429 you're making a larger attack surface. 498 00:17:35,430 --> 00:17:37,799 You are creating additional possibilities 499 00:17:37,800 --> 00:17:39,000 for vulnerabilities. 500 00:17:40,150 --> 00:17:42,269 Also, with any case of of an escrow 501 00:17:42,270 --> 00:17:44,759 agent, you have some questions 502 00:17:44,760 --> 00:17:46,619 arise. Who would be that escrow? 503 00:17:46,620 --> 00:17:48,839 Would the government be in escrow 504 00:17:48,840 --> 00:17:50,609 then? If so, how do you pick which 505 00:17:50,610 --> 00:17:52,439 government or which government gets 506 00:17:52,440 --> 00:17:54,029 access to the escrow? 507 00:17:54,030 --> 00:17:56,789 Should the provider be the escrow? 508 00:17:56,790 --> 00:17:59,369 Should we have a trusted third party? 509 00:17:59,370 --> 00:18:01,469 And for any of the escrows, they're 510 00:18:01,470 --> 00:18:03,479 often not in a position to raise the 511 00:18:03,480 --> 00:18:05,579 important concerns that we might want 512 00:18:05,580 --> 00:18:07,679 to be have handled. For example, if we've 513 00:18:07,680 --> 00:18:09,719 decided that certain communications are 514 00:18:09,720 --> 00:18:11,219 privileged like an attorney client 515 00:18:11,220 --> 00:18:13,619 communication, none of these escrows 516 00:18:13,620 --> 00:18:15,779 are in a position to know 517 00:18:15,780 --> 00:18:17,159 that this is a privileged communication 518 00:18:17,160 --> 00:18:19,289 to raise that concern and 519 00:18:20,760 --> 00:18:22,919 failed, you know, resist providing and 520 00:18:22,920 --> 00:18:24,690 challenge the acquisition. 521 00:18:25,800 --> 00:18:28,439 If you have an escrow, you also introduce 522 00:18:28,440 --> 00:18:29,849 insider risk. 523 00:18:29,850 --> 00:18:31,469 The possibility that someone who's 524 00:18:31,470 --> 00:18:33,269 working at the escrow agent is familiar 525 00:18:33,270 --> 00:18:35,369 with its practices can do something 526 00:18:35,370 --> 00:18:37,259 that would make it appear like they have 527 00:18:37,260 --> 00:18:39,389 a legitimate request that 528 00:18:39,390 --> 00:18:41,579 has come through the proper channels and 529 00:18:41,580 --> 00:18:44,459 then use that to to gain access 530 00:18:44,460 --> 00:18:46,529 to the previously 531 00:18:46,530 --> 00:18:48,479 private communications. 532 00:18:48,480 --> 00:18:50,669 And we have a bit of a history to look 533 00:18:50,670 --> 00:18:53,309 back on why this can be problematic 534 00:18:53,310 --> 00:18:55,199 with law enforcement access points. 535 00:18:55,200 --> 00:18:57,419 So there have been a number 536 00:18:57,420 --> 00:18:59,669 of attacks on law enforcement 537 00:18:59,670 --> 00:19:01,079 access points. 538 00:19:01,080 --> 00:19:03,389 Some become a tempting target 539 00:19:03,390 --> 00:19:05,189 for criminals, for state sponsored 540 00:19:05,190 --> 00:19:06,389 attackers. 541 00:19:06,390 --> 00:19:09,119 We look at the example of the wiretapping 542 00:19:09,120 --> 00:19:11,519 scandal where it was a law enforcement 543 00:19:11,520 --> 00:19:13,799 access point, that once 544 00:19:13,800 --> 00:19:15,929 you get access to the access point, 545 00:19:15,930 --> 00:19:17,519 you can use its properties and its 546 00:19:17,520 --> 00:19:20,219 permissions to get a much wider access. 547 00:19:20,220 --> 00:19:22,679 Similarly, China attacked Google's 548 00:19:22,680 --> 00:19:24,839 law enforcement access point, using that 549 00:19:24,840 --> 00:19:26,730 to get information about dissidents. 550 00:19:27,960 --> 00:19:30,059 And so if you create an 551 00:19:30,060 --> 00:19:32,249 escrow agent, if you create a method for 552 00:19:32,250 --> 00:19:34,049 law enforcement to get access to it, you 553 00:19:34,050 --> 00:19:36,149 create a very tempting target that will 554 00:19:36,150 --> 00:19:38,009 be certainly in the sights of any 555 00:19:38,010 --> 00:19:39,389 attackers. 556 00:19:39,390 --> 00:19:41,069 So a lot of these arguments have been 557 00:19:41,070 --> 00:19:43,409 raised. And what we heard back from 558 00:19:43,410 --> 00:19:45,329 from some people in the policymaker 559 00:19:45,330 --> 00:19:47,549 things is, OK, we see that 560 00:19:47,550 --> 00:19:49,410 you have a problem with back doors. 561 00:19:51,420 --> 00:19:53,669 Maybe we should just rename it so 562 00:19:53,670 --> 00:19:54,779 we don't want a back door. 563 00:19:54,780 --> 00:19:55,829 We want a front door. 564 00:19:56,910 --> 00:19:57,839 Or, you know, sometimes people have 565 00:19:57,840 --> 00:20:00,449 talked about side door, trap door. 566 00:20:01,470 --> 00:20:03,719 And this is very sort of 567 00:20:03,720 --> 00:20:06,659 typical of politicians 568 00:20:06,660 --> 00:20:09,209 is that, you know, I think the problem is 569 00:20:09,210 --> 00:20:11,639 the rhetoric, the name 570 00:20:11,640 --> 00:20:14,519 that we can you know, The Washington Post 571 00:20:14,520 --> 00:20:16,499 also saying, well, we understand that a 572 00:20:16,500 --> 00:20:18,779 backdoor can be exploited by 573 00:20:18,780 --> 00:20:20,219 bad guys. 574 00:20:20,220 --> 00:20:22,409 So instead, how about a secure 575 00:20:22,410 --> 00:20:23,410 golden key? 576 00:20:24,120 --> 00:20:26,849 And, you know, this this was widely, 577 00:20:26,850 --> 00:20:29,219 widely criticized and ridiculed, 578 00:20:29,220 --> 00:20:31,139 as it should have been. 579 00:20:31,140 --> 00:20:33,389 But it reflects an attitude 580 00:20:33,390 --> 00:20:35,459 of we'll just 581 00:20:35,460 --> 00:20:37,499 say what we want, which is something 582 00:20:37,500 --> 00:20:39,629 magical, a secure golden key, 583 00:20:39,630 --> 00:20:41,939 and then tossed the problem back 584 00:20:41,940 --> 00:20:43,649 to Technologist's to. 585 00:20:43,650 --> 00:20:45,869 Trying to solve and maybe even legislate, 586 00:20:45,870 --> 00:20:48,089 so they have to solve it 587 00:20:48,090 --> 00:20:50,309 and sort of reflecting an attitude 588 00:20:50,310 --> 00:20:51,929 that this is all they use, the term 589 00:20:51,930 --> 00:20:54,389 wizardry, it is magical wizardry 590 00:20:54,390 --> 00:20:56,099 that people can do and come up with 591 00:20:56,100 --> 00:20:57,240 magical solutions. 592 00:20:59,910 --> 00:21:01,980 So while this is going on, it was still 593 00:21:03,060 --> 00:21:04,829 they're having some troubles that, you 594 00:21:04,830 --> 00:21:06,659 know, these proposals will were ridiculed 595 00:21:06,660 --> 00:21:08,519 over a lot of people who are pointing out 596 00:21:08,520 --> 00:21:11,699 the many flaws with back doors. 597 00:21:11,700 --> 00:21:13,769 But they held 598 00:21:13,770 --> 00:21:15,659 out an interesting hope. 599 00:21:15,660 --> 00:21:17,489 And this is a quote from the general 600 00:21:17,490 --> 00:21:19,169 counsel of the office of the director of 601 00:21:19,170 --> 00:21:21,539 National Intelligence, the 602 00:21:21,540 --> 00:21:24,809 oversight agency that 603 00:21:24,810 --> 00:21:26,129 looks over the U.S. 604 00:21:26,130 --> 00:21:27,239 intelligence community. 605 00:21:28,570 --> 00:21:30,689 Well, what if there was a terrorist 606 00:21:30,690 --> 00:21:32,579 attack or criminal event that could be 607 00:21:32,580 --> 00:21:34,709 shown to have where encryption had 608 00:21:34,710 --> 00:21:36,059 hindered law enforcement? 609 00:21:36,060 --> 00:21:38,039 That's the real thing that we need to to 610 00:21:38,040 --> 00:21:40,559 turn this discussion around. 611 00:21:40,560 --> 00:21:42,629 So we had 612 00:21:42,630 --> 00:21:44,699 a number of attacks in 613 00:21:44,700 --> 00:21:46,289 this last term, tragic attacks. 614 00:21:47,370 --> 00:21:49,619 And but what what do you do if 615 00:21:49,620 --> 00:21:51,719 you have these things? But Krypto wasn't 616 00:21:51,720 --> 00:21:52,679 involved. 617 00:21:52,680 --> 00:21:55,889 Well, you suggest that it was anyway. 618 00:21:55,890 --> 00:21:58,079 So we have the the prescient comment here 619 00:21:58,080 --> 00:22:00,329 from Michael Morell, the former 620 00:22:00,330 --> 00:22:02,429 CIA deputy director. 621 00:22:02,430 --> 00:22:04,529 We don't know anything, but I bet you the 622 00:22:04,530 --> 00:22:05,639 encryption was involved. 623 00:22:05,640 --> 00:22:07,739 And then he proceeded to talk about how 624 00:22:07,740 --> 00:22:10,709 how encryption was very, very dangerous. 625 00:22:10,710 --> 00:22:12,869 So I following the attacks in 626 00:22:12,870 --> 00:22:15,059 Paris, attacks in San Bernardino, 627 00:22:15,060 --> 00:22:17,129 there were a lot of talking heads who 628 00:22:17,130 --> 00:22:19,619 are going on the television shows, 629 00:22:19,620 --> 00:22:22,199 the public policy debates 630 00:22:22,200 --> 00:22:24,359 and trying to say that encryption 631 00:22:24,360 --> 00:22:27,029 was at fault when in fact, 632 00:22:27,030 --> 00:22:29,430 the Paris attackers use plain text 633 00:22:30,660 --> 00:22:32,309 text messages. 634 00:22:32,310 --> 00:22:34,019 The San Bernardino attackers actually 635 00:22:34,020 --> 00:22:35,849 mostly they communicated just directly in 636 00:22:35,850 --> 00:22:38,219 person, but 637 00:22:38,220 --> 00:22:40,949 they also use some direct messages 638 00:22:40,950 --> 00:22:42,119 that were not encrypted. 639 00:22:44,810 --> 00:22:47,149 And they're also trying to take this time 640 00:22:47,150 --> 00:22:49,249 to demonize encryption, to 641 00:22:49,250 --> 00:22:51,589 try and get people with fear, 642 00:22:51,590 --> 00:22:53,689 uncertainty and doubt, the home 643 00:22:53,690 --> 00:22:55,919 secretary may brings up what? 644 00:22:55,920 --> 00:22:57,979 What? Back in the 90s, we had called 645 00:22:57,980 --> 00:23:00,379 the, you know, four horsemen of the info 646 00:23:00,380 --> 00:23:02,719 apocalypse, talking about 647 00:23:02,720 --> 00:23:04,939 child pornographers, crime, 648 00:23:04,940 --> 00:23:08,149 drugs, terrorists. 649 00:23:08,150 --> 00:23:10,579 The Senator Feinstein, who is on the 650 00:23:10,580 --> 00:23:12,769 U.S. Intelligence Committee, went even 651 00:23:12,770 --> 00:23:14,989 a little further, basically saying 652 00:23:14,990 --> 00:23:17,119 that encryption will be used to 653 00:23:17,120 --> 00:23:18,319 behead children, 654 00:23:19,610 --> 00:23:20,990 which, you know, I think this is 655 00:23:22,010 --> 00:23:24,019 taking it a little far, a little fun. 656 00:23:28,690 --> 00:23:30,789 So with this in 657 00:23:30,790 --> 00:23:32,649 this atmosphere, we're starting to see 658 00:23:32,650 --> 00:23:35,259 some proposals come out of many countries 659 00:23:35,260 --> 00:23:37,599 to try to limit encryption, sometimes 660 00:23:37,600 --> 00:23:39,879 to they're talking about mandating a back 661 00:23:39,880 --> 00:23:42,369 door, sometimes mandating access 662 00:23:42,370 --> 00:23:44,589 to plain text without sort of specifying 663 00:23:44,590 --> 00:23:46,689 the the manner of the access 664 00:23:46,690 --> 00:23:48,939 or putting laws that would help 665 00:23:48,940 --> 00:23:51,069 in danger. Encryption. 666 00:23:51,070 --> 00:23:52,479 Probably the most prominent of these 667 00:23:52,480 --> 00:23:54,639 right now is what's known as the 668 00:23:54,640 --> 00:23:56,979 snoopers charter, the Investigatory 669 00:23:56,980 --> 00:24:00,099 Powers Act out of the United Kingdom. 670 00:24:00,100 --> 00:24:02,199 Now, this is an interesting proposal 671 00:24:02,200 --> 00:24:05,019 because it is not just aimed at 672 00:24:05,020 --> 00:24:06,909 telecommunication companies in the United 673 00:24:06,910 --> 00:24:09,099 Kingdom. It reports to regulate 674 00:24:09,100 --> 00:24:11,349 companies all over the world. 675 00:24:11,350 --> 00:24:12,849 Now, this raises some questions. 676 00:24:12,850 --> 00:24:14,929 You know, if you were had a company who 677 00:24:14,930 --> 00:24:16,869 did no business in the United Kingdom, 678 00:24:16,870 --> 00:24:18,639 you were completely separate from it. 679 00:24:18,640 --> 00:24:20,019 You know, how would they be able to 680 00:24:20,020 --> 00:24:21,009 enforce that? 681 00:24:21,010 --> 00:24:22,959 But for a large company that has offices 682 00:24:22,960 --> 00:24:24,189 all over the world, many of them will 683 00:24:24,190 --> 00:24:26,649 have offices or people on the ground 684 00:24:26,650 --> 00:24:27,789 within the U.K. 685 00:24:27,790 --> 00:24:30,189 that would be subject to the UK's 686 00:24:30,190 --> 00:24:31,629 jurisdiction. 687 00:24:31,630 --> 00:24:33,999 And this can be a 688 00:24:34,000 --> 00:24:35,319 tool that the U.K. 689 00:24:35,320 --> 00:24:37,419 can use to try and enforce this against 690 00:24:37,420 --> 00:24:38,619 them. 691 00:24:38,620 --> 00:24:40,719 And it's a long bill as 299 692 00:24:40,720 --> 00:24:41,619 pages. 693 00:24:41,620 --> 00:24:43,209 A lot of detail is actually a lot of 694 00:24:43,210 --> 00:24:45,009 problems with the bill beyond the 695 00:24:45,010 --> 00:24:46,479 encryption issue. But today we want to 696 00:24:46,480 --> 00:24:47,649 focus on that. 697 00:24:47,650 --> 00:24:49,809 And in Section 189, 698 00:24:49,810 --> 00:24:51,879 they have a provision dealing with 699 00:24:51,880 --> 00:24:53,619 electronic protection. 700 00:24:53,620 --> 00:24:55,839 This is a term which we believe they 701 00:24:55,840 --> 00:24:57,220 mean encryption. 702 00:24:58,900 --> 00:25:00,969 And it is saying that 703 00:25:00,970 --> 00:25:03,069 if the provider put 704 00:25:03,070 --> 00:25:05,289 on the electronic protection, it 705 00:25:05,290 --> 00:25:08,229 may be obligated to take it off. 706 00:25:08,230 --> 00:25:10,329 Now, this could 707 00:25:10,330 --> 00:25:12,169 be interpreted a variety of ways. 708 00:25:12,170 --> 00:25:13,839 It will be interesting to see how it 709 00:25:13,840 --> 00:25:15,939 eventually gets interpreted, 710 00:25:15,940 --> 00:25:17,709 but it could be interpreted to require 711 00:25:17,710 --> 00:25:20,709 weakening encryption, holding a key 712 00:25:20,710 --> 00:25:22,479 banning end to end encryption if it is 713 00:25:22,480 --> 00:25:25,029 not possible to do this. 714 00:25:25,030 --> 00:25:26,319 And one of the things that's interesting 715 00:25:26,320 --> 00:25:28,449 about it, so it has some language there 716 00:25:28,450 --> 00:25:29,450 saying that 717 00:25:30,520 --> 00:25:32,649 only if it's practicable, 718 00:25:32,650 --> 00:25:34,239 only if it's, you know, feasible, 719 00:25:34,240 --> 00:25:35,619 reasonable to do this. 720 00:25:35,620 --> 00:25:37,749 But this is done in the home secretary 721 00:25:37,750 --> 00:25:39,789 determination. So it won't be the 722 00:25:39,790 --> 00:25:41,349 technologies that would be able to 723 00:25:41,350 --> 00:25:43,419 determine whether it was practical to 724 00:25:43,420 --> 00:25:45,549 be able to remove the encryption, but 725 00:25:45,550 --> 00:25:47,199 rather the home secretary say, I think 726 00:25:47,200 --> 00:25:48,549 that's practical. And so now you're 727 00:25:48,550 --> 00:25:50,739 obligated to do that. 728 00:25:50,740 --> 00:25:52,119 So the latest version of the 729 00:25:52,120 --> 00:25:53,889 Investigatory Powers Act came out in 730 00:25:53,890 --> 00:25:54,909 November. 731 00:25:54,910 --> 00:25:56,979 It is currently before a committee in 732 00:25:56,980 --> 00:25:59,229 the U.K. parliament, which is accepting 733 00:25:59,230 --> 00:26:00,999 comments right now. 734 00:26:01,000 --> 00:26:03,249 A number of organizations, EFF and many 735 00:26:03,250 --> 00:26:05,199 other civil society organizations, have 736 00:26:05,200 --> 00:26:06,279 submitted comment. 737 00:26:06,280 --> 00:26:07,779 A lot of industry have submitted 738 00:26:07,780 --> 00:26:08,859 comments. 739 00:26:08,860 --> 00:26:11,049 Apple in particular made some news with 740 00:26:11,050 --> 00:26:13,209 its comments with a strong 741 00:26:13,210 --> 00:26:15,489 defense of privacy 742 00:26:15,490 --> 00:26:17,649 and warning against back 743 00:26:17,650 --> 00:26:19,389 doors, saying that a key left under the 744 00:26:19,390 --> 00:26:21,489 doormat would not just be there for 745 00:26:21,490 --> 00:26:23,689 the good guys, but recognizing 746 00:26:23,690 --> 00:26:25,930 the bad guys would find it, too. 747 00:26:27,310 --> 00:26:28,419 There been a couple of other interesting 748 00:26:28,420 --> 00:26:30,399 ones around the world. 749 00:26:30,400 --> 00:26:32,919 Australia has a Defense Trade 750 00:26:32,920 --> 00:26:35,049 Controls Act that has some 751 00:26:35,050 --> 00:26:37,059 interesting language talking about the 752 00:26:37,060 --> 00:26:39,069 intangible supply. 753 00:26:39,070 --> 00:26:41,859 So, you know, this is a, 754 00:26:41,860 --> 00:26:43,599 you know, munitions control act and 755 00:26:43,600 --> 00:26:44,650 export control. 756 00:26:45,850 --> 00:26:47,649 And, you know, it's pretty obvious when 757 00:26:47,650 --> 00:26:49,509 you are supplying a tank, whether, you 758 00:26:49,510 --> 00:26:52,149 know, supply or not, it's not very 759 00:26:52,150 --> 00:26:53,529 hard to figure out. 760 00:26:53,530 --> 00:26:55,359 But when you're talking about encryption, 761 00:26:55,360 --> 00:26:57,489 it gets a little bit weird because 762 00:26:57,490 --> 00:26:58,839 you know what? If you explain the 763 00:26:58,840 --> 00:27:00,849 algorithm to somebody, what if you showed 764 00:27:00,850 --> 00:27:02,289 them a copy of the algorithm? 765 00:27:02,290 --> 00:27:03,579 Is that exporting? 766 00:27:03,580 --> 00:27:05,489 Is that supplying the technology? 767 00:27:07,090 --> 00:27:08,799 The Department of Defense in Australia 768 00:27:08,800 --> 00:27:11,259 has suggested that merely explaining 769 00:27:11,260 --> 00:27:13,719 an algorithm could be intangible 770 00:27:13,720 --> 00:27:15,879 supply come within the zone 771 00:27:15,880 --> 00:27:16,880 of this act. 772 00:27:17,890 --> 00:27:19,659 Now, it remains to be seen whether that 773 00:27:19,660 --> 00:27:21,969 interpretation will be will be upheld in 774 00:27:21,970 --> 00:27:22,869 the courts. 775 00:27:22,870 --> 00:27:24,869 But what this means is that there is a 776 00:27:26,040 --> 00:27:28,359 a threat, a possibility 777 00:27:28,360 --> 00:27:29,979 that somebody who is doing ordinary 778 00:27:29,980 --> 00:27:32,289 teaching and research activities could 779 00:27:32,290 --> 00:27:34,509 be subject to these controls with very 780 00:27:34,510 --> 00:27:35,409 severe penalties. 781 00:27:35,410 --> 00:27:37,269 And that provides a chilling effect for 782 00:27:37,270 --> 00:27:39,069 people trying to do things which are very 783 00:27:39,070 --> 00:27:41,229 important for our security by doing 784 00:27:41,230 --> 00:27:43,509 the research, doing the teaching. 785 00:27:43,510 --> 00:27:45,729 So a lot of people have 786 00:27:45,730 --> 00:27:47,649 signed a petition against hundreds of 787 00:27:47,650 --> 00:27:49,809 experts have explained why this 788 00:27:49,810 --> 00:27:51,359 is a very, very bad idea. 789 00:27:52,810 --> 00:27:55,479 India also came out this September 790 00:27:55,480 --> 00:27:57,849 with a draft national encryption 791 00:27:57,850 --> 00:27:59,379 policy. 792 00:27:59,380 --> 00:28:01,509 This was a sort of particularly 793 00:28:01,510 --> 00:28:02,859 bad one. 794 00:28:02,860 --> 00:28:04,749 It said that everyone was required to 795 00:28:04,750 --> 00:28:06,879 store a plain text copy of 796 00:28:06,880 --> 00:28:08,859 their encrypted messages, keep that 797 00:28:08,860 --> 00:28:11,109 around and then be able to hand 798 00:28:11,110 --> 00:28:13,209 it over upon request. 799 00:28:13,210 --> 00:28:15,339 This was widely condemned and ridiculed 800 00:28:15,340 --> 00:28:17,589 and eventually India 801 00:28:17,590 --> 00:28:19,209 withdrew this. 802 00:28:19,210 --> 00:28:20,409 They're going to go back to the drawing 803 00:28:20,410 --> 00:28:21,939 boards, but we'll have to keep an eye out 804 00:28:21,940 --> 00:28:23,950 to see what they come up with next. 805 00:28:25,150 --> 00:28:28,089 Very recently, actually, on Sunday 806 00:28:28,090 --> 00:28:30,369 of, I guess four days 807 00:28:30,370 --> 00:28:32,889 ago, China passed a new 808 00:28:32,890 --> 00:28:35,109 anti-terrorism law. 809 00:28:35,110 --> 00:28:37,599 In a draft version, there was a provision 810 00:28:37,600 --> 00:28:39,669 that would require technology 811 00:28:39,670 --> 00:28:41,559 companies to hand over their encryption 812 00:28:41,560 --> 00:28:42,489 codes. 813 00:28:42,490 --> 00:28:44,559 The final version did remove that, 814 00:28:44,560 --> 00:28:46,509 but it still contains some very dangerous 815 00:28:46,510 --> 00:28:49,269 language, saying that the companies 816 00:28:49,270 --> 00:28:51,519 are required to provide technical 817 00:28:51,520 --> 00:28:53,619 interfaces, decryption and 818 00:28:53,620 --> 00:28:55,269 other technical support. 819 00:28:55,270 --> 00:28:57,759 So, again, this could be interpreted 820 00:28:57,760 --> 00:28:59,859 to mean that they have to find a way 821 00:28:59,860 --> 00:29:01,179 to be able to decrypt it, which means 822 00:29:01,180 --> 00:29:03,339 having some weakness, some backdoor, 823 00:29:03,340 --> 00:29:04,929 some additional key that would enable 824 00:29:04,930 --> 00:29:06,789 them to do so. 825 00:29:06,790 --> 00:29:08,109 And one of the things that actually also 826 00:29:08,110 --> 00:29:10,899 came up from China's anti-terrorism 827 00:29:10,900 --> 00:29:13,269 law is that 828 00:29:13,270 --> 00:29:15,519 when discussing it, they were pointing 829 00:29:15,520 --> 00:29:17,649 to other countries who 830 00:29:17,650 --> 00:29:19,929 have put in or proposed similar 831 00:29:19,930 --> 00:29:21,339 things so that 832 00:29:22,420 --> 00:29:24,849 in when when the governments 833 00:29:24,850 --> 00:29:27,009 around the world put forward proposals to 834 00:29:27,010 --> 00:29:29,229 create back doors to to 835 00:29:29,230 --> 00:29:31,209 suggest that encryption is a problem and 836 00:29:31,210 --> 00:29:32,919 they mean it, say, well, we're trying to 837 00:29:32,920 --> 00:29:35,019 make it so that it's safe for, 838 00:29:35,020 --> 00:29:37,779 you know, Western democracy. 839 00:29:37,780 --> 00:29:40,479 At the same time, regimes which 840 00:29:40,480 --> 00:29:42,669 are much more 841 00:29:42,670 --> 00:29:44,889 totalitarian can use the same 842 00:29:44,890 --> 00:29:46,779 language, use the same rhetoric to 843 00:29:46,780 --> 00:29:49,029 justify their own attacks and 844 00:29:49,030 --> 00:29:51,069 their own attempts to make it so people 845 00:29:51,070 --> 00:29:52,719 can't have secure and private 846 00:29:52,720 --> 00:29:53,720 communications 847 00:29:55,090 --> 00:29:56,299 in the United States. 848 00:29:56,300 --> 00:29:57,430 There is not yet 849 00:29:58,540 --> 00:30:00,669 legislation to mandate 850 00:30:00,670 --> 00:30:02,469 back doors and mandate access to 851 00:30:02,470 --> 00:30:04,059 encryption. 852 00:30:04,060 --> 00:30:06,279 President Obama has said that he 853 00:30:06,280 --> 00:30:08,649 will not, for now, call 854 00:30:08,650 --> 00:30:10,509 for this legislation. 855 00:30:10,510 --> 00:30:13,299 This is for now is somewhat worrisome. 856 00:30:13,300 --> 00:30:15,409 And we organized 857 00:30:15,410 --> 00:30:17,199 with with several other civil society 858 00:30:17,200 --> 00:30:19,539 groups a petition asking 859 00:30:19,540 --> 00:30:21,789 him to support strong 860 00:30:21,790 --> 00:30:24,009 encryption after he had put 861 00:30:24,010 --> 00:30:25,629 out this final statement to make a more 862 00:30:25,630 --> 00:30:27,639 clear statement that they would put that 863 00:30:27,640 --> 00:30:29,529 aside and say that, no, they weren't 864 00:30:29,530 --> 00:30:31,629 going to ask for back doors, not just for 865 00:30:31,630 --> 00:30:34,339 now, but but forever. 866 00:30:34,340 --> 00:30:36,459 And we have not yet 867 00:30:36,460 --> 00:30:38,679 gotten the full response to that. 868 00:30:38,680 --> 00:30:41,049 So hopefully that will come soon and 869 00:30:41,050 --> 00:30:42,969 it will be a clearer statement. 870 00:30:42,970 --> 00:30:45,219 But right now, we're sort of in a waiting 871 00:30:45,220 --> 00:30:47,139 game to see whether it will go forward. 872 00:30:47,140 --> 00:30:49,299 And the rumor has it that the Senate 873 00:30:49,300 --> 00:30:51,219 Intelligence Committee is going to be 874 00:30:51,220 --> 00:30:53,589 proposing a bill in the coming 875 00:30:53,590 --> 00:30:55,599 spring. Now, if they do propose that, of 876 00:30:55,600 --> 00:30:57,789 course, we will fight against that bill, 877 00:30:57,790 --> 00:30:59,139 try and make sure that it doesn't get 878 00:30:59,140 --> 00:31:01,509 passed and if it gets passed, 879 00:31:01,510 --> 00:31:03,759 move to get it thrown 880 00:31:03,760 --> 00:31:06,849 out in the courts as as unconstitutional. 881 00:31:08,630 --> 00:31:10,369 I want to take a brief moment to talk 882 00:31:10,370 --> 00:31:12,559 about the Trans-Pacific Partnership, 883 00:31:12,560 --> 00:31:14,899 this is a international trade 884 00:31:14,900 --> 00:31:17,029 agreement that has been negotiated, 885 00:31:17,030 --> 00:31:19,729 is now being considered. 886 00:31:19,730 --> 00:31:22,069 And some people look through this 887 00:31:22,070 --> 00:31:23,899 lengthy agreement once it became public 888 00:31:23,900 --> 00:31:25,939 and noted that there was a little bit on 889 00:31:25,940 --> 00:31:27,769 encryption and there was a question, 890 00:31:27,770 --> 00:31:29,629 well, was there actually some good news 891 00:31:29,630 --> 00:31:31,759 in this? Otherwise, the TPP is fairly 892 00:31:31,760 --> 00:31:33,319 terrible. 893 00:31:33,320 --> 00:31:35,839 And it turns out in the final analysis, 894 00:31:35,840 --> 00:31:37,999 no, there was a provision in there or is 895 00:31:38,000 --> 00:31:39,709 a provision in there saying that a 896 00:31:39,710 --> 00:31:41,719 provider may not be compelled to give a 897 00:31:41,720 --> 00:31:44,269 key, but only as a condition 898 00:31:44,270 --> 00:31:46,129 of sale. So it's silent as to whether 899 00:31:46,130 --> 00:31:47,899 they could be compelled to give a key 900 00:31:47,900 --> 00:31:49,639 under other conditions. 901 00:31:49,640 --> 00:31:51,589 And there's also a provision in there 902 00:31:51,590 --> 00:31:53,839 that said that the provider can be 903 00:31:53,840 --> 00:31:56,269 required to give decrypted content, 904 00:31:56,270 --> 00:31:57,829 which still has a lot of the same 905 00:31:57,830 --> 00:32:00,019 problems that 906 00:32:00,020 --> 00:32:01,339 we have been we have been fighting 907 00:32:01,340 --> 00:32:02,340 against. 908 00:32:04,100 --> 00:32:05,959 So at the same time as this legislation 909 00:32:05,960 --> 00:32:08,389 is going on, the public policy debates 910 00:32:08,390 --> 00:32:10,489 in the background, they're also 911 00:32:10,490 --> 00:32:12,559 technical attempts to try and 912 00:32:12,560 --> 00:32:14,629 reduce the effectiveness of encryption, 913 00:32:14,630 --> 00:32:16,189 make it harder to use harder use 914 00:32:16,190 --> 00:32:17,190 securely. 915 00:32:17,960 --> 00:32:19,459 One of the more prominent ones that came 916 00:32:19,460 --> 00:32:21,559 out of the Snowden documents there was 917 00:32:21,560 --> 00:32:23,299 routing around encryption. 918 00:32:23,300 --> 00:32:25,399 We have here the illustration 919 00:32:25,400 --> 00:32:28,399 where they looked at the communications 920 00:32:28,400 --> 00:32:30,649 between data centers where 921 00:32:30,650 --> 00:32:32,869 Google added and removed SSL 922 00:32:32,870 --> 00:32:34,429 at a particular point and they were able 923 00:32:34,430 --> 00:32:36,469 to get in there and get the unencrypted 924 00:32:36,470 --> 00:32:37,639 communications. 925 00:32:37,640 --> 00:32:39,739 Now, when that came out, this 926 00:32:39,740 --> 00:32:41,929 caused Google and 927 00:32:41,930 --> 00:32:43,699 many other companies to beef up their 928 00:32:43,700 --> 00:32:45,769 encryption and encrypt between the data 929 00:32:45,770 --> 00:32:47,989 centers. But this is still 930 00:32:47,990 --> 00:32:49,609 people are still looking for it. 931 00:32:49,610 --> 00:32:52,429 And you can rest assured that the 932 00:32:52,430 --> 00:32:54,529 intelligence agencies are trying to find 933 00:32:54,530 --> 00:32:56,629 places where they can just go find a 934 00:32:56,630 --> 00:32:59,059 gap in encryption and get the information 935 00:32:59,060 --> 00:33:01,189 there. They're also working on breaking 936 00:33:01,190 --> 00:33:03,229 encryption and certain vulnerabilities 937 00:33:03,230 --> 00:33:04,459 and putting on malware. 938 00:33:04,460 --> 00:33:06,019 Go through a couple of these. 939 00:33:06,020 --> 00:33:07,879 One, as we know from the Snowden 940 00:33:07,880 --> 00:33:10,669 documents about the Bull Run program. 941 00:33:10,670 --> 00:33:12,259 Two hundred and fifty million dollar a 942 00:33:12,260 --> 00:33:15,109 year budget to insert vulnerabilities, 943 00:33:15,110 --> 00:33:17,569 to influence policies, to participate 944 00:33:17,570 --> 00:33:19,669 in cryptography 945 00:33:19,670 --> 00:33:21,949 discussions, to try to weaken and 946 00:33:21,950 --> 00:33:23,659 influence the standards that are being 947 00:33:23,660 --> 00:33:24,769 used. 948 00:33:24,770 --> 00:33:27,019 And we saw a an example 949 00:33:27,020 --> 00:33:29,749 of that with the Orissa's 950 00:33:29,750 --> 00:33:31,969 use of the dual Essy 951 00:33:31,970 --> 00:33:33,619 D.R big 952 00:33:34,670 --> 00:33:36,349 encryption standard or the random number 953 00:33:36,350 --> 00:33:38,029 generator. 954 00:33:38,030 --> 00:33:40,309 This had a flaw 955 00:33:40,310 --> 00:33:42,379 that if there was a I'm 956 00:33:42,380 --> 00:33:43,759 only going to go over the technology of 957 00:33:43,760 --> 00:33:45,049 this very briefly. We don't have much 958 00:33:45,050 --> 00:33:47,929 time. But there was a constant queue 959 00:33:47,930 --> 00:33:49,669 that if you if you knew what it was, if 960 00:33:49,670 --> 00:33:51,799 you made it special, 961 00:33:51,800 --> 00:33:54,439 would operate to significantly 962 00:33:54,440 --> 00:33:56,689 reduce the complexity of an attack and 963 00:33:56,690 --> 00:33:58,969 basically be able to backdoor the random 964 00:33:58,970 --> 00:34:01,129 number generator and therefore be able 965 00:34:01,130 --> 00:34:03,259 to more easily get 966 00:34:03,260 --> 00:34:04,519 access to encryption. 967 00:34:05,570 --> 00:34:07,849 And the NSA paid RSA ten million 968 00:34:07,850 --> 00:34:09,980 dollars to make it the default. 969 00:34:12,230 --> 00:34:13,249 And we had. 970 00:34:13,250 --> 00:34:16,039 So this was known to have a potential 971 00:34:16,040 --> 00:34:18,499 back doors early as 2007 972 00:34:19,520 --> 00:34:21,468 after some of the Snowden documents came 973 00:34:21,469 --> 00:34:23,569 out. After we got some reports 974 00:34:23,570 --> 00:34:25,099 about the 10 million dollar payment, 975 00:34:25,100 --> 00:34:27,319 people looked at it very deeply 976 00:34:27,320 --> 00:34:29,718 and showed how it could be used and prove 977 00:34:29,719 --> 00:34:32,238 that it could be used for this backdoor. 978 00:34:32,239 --> 00:34:34,939 But very recently, we actually got a 979 00:34:34,940 --> 00:34:37,549 an interesting example of this backdoor 980 00:34:37,550 --> 00:34:39,379 in operation, but it's a very curious 981 00:34:39,380 --> 00:34:40,399 case. 982 00:34:40,400 --> 00:34:42,649 So Juniper uses an operating 983 00:34:42,650 --> 00:34:44,869 system for some of its routing software 984 00:34:44,870 --> 00:34:47,988 that used this dual AC 985 00:34:47,989 --> 00:34:50,479 program, but it didn't use the 986 00:34:50,480 --> 00:34:52,638 CPU that had been suggested by 987 00:34:52,639 --> 00:34:54,379 the NSA. 988 00:34:54,380 --> 00:34:56,449 They use a new and alternative cue. 989 00:34:57,620 --> 00:34:59,119 So perhaps this means that, you know, 990 00:34:59,120 --> 00:35:00,739 there was a back door, but but they 991 00:35:00,740 --> 00:35:01,819 changed the locks. 992 00:35:03,200 --> 00:35:05,389 And in addition, the output from dual 993 00:35:05,390 --> 00:35:07,999 HESI was passed through a second stronger 994 00:35:08,000 --> 00:35:10,339 random number generator, which, 995 00:35:10,340 --> 00:35:13,099 you know, you may 996 00:35:13,100 --> 00:35:15,589 have made it so that 997 00:35:15,590 --> 00:35:18,229 it was it was not a big deal, that 998 00:35:18,230 --> 00:35:19,849 it would be put through the stronger one 999 00:35:19,850 --> 00:35:21,949 and made it made as strong as that. 1000 00:35:21,950 --> 00:35:23,819 However, people discover by looking at 1001 00:35:23,820 --> 00:35:26,209 the at the code that there was 1002 00:35:26,210 --> 00:35:28,519 a portion that was looking at the raw 1003 00:35:28,520 --> 00:35:30,649 output, which is the purpose of which 1004 00:35:30,650 --> 00:35:32,509 would give them these 32 byte to be 1005 00:35:32,510 --> 00:35:34,639 necessary in order to 1006 00:35:34,640 --> 00:35:36,949 passively break 1007 00:35:36,950 --> 00:35:37,950 VPN. 1008 00:35:39,230 --> 00:35:40,999 And in addition, there was a second floor 1009 00:35:41,000 --> 00:35:43,039 that came out was that somebody had hard 1010 00:35:43,040 --> 00:35:45,169 coded a password for 1011 00:35:45,170 --> 00:35:47,869 S.H. and Telnet into the program. 1012 00:35:47,870 --> 00:35:49,309 These look to be perhaps different 1013 00:35:49,310 --> 00:35:51,649 attacks there. There one is good for 1014 00:35:51,650 --> 00:35:53,179 passive collection of VPN. 1015 00:35:53,180 --> 00:35:55,309 The other is good for going in and 1016 00:35:55,310 --> 00:35:58,159 owning a particular router. 1017 00:35:58,160 --> 00:36:00,349 But what is very curious about this 1018 00:36:00,350 --> 00:36:02,749 is what exactly happened 1019 00:36:02,750 --> 00:36:04,999 here. It seems like someone 1020 00:36:05,000 --> 00:36:07,099 came in, took advantage of 1021 00:36:07,100 --> 00:36:09,169 the back door that the NSA had 1022 00:36:09,170 --> 00:36:11,479 created through the Dualeh Sea 1023 00:36:11,480 --> 00:36:12,979 and used it for their own devices. 1024 00:36:12,980 --> 00:36:15,169 This is illustrating what are the great 1025 00:36:15,170 --> 00:36:17,179 dangers of putting in a backdoor. 1026 00:36:17,180 --> 00:36:19,339 They depend on the notion that only 1027 00:36:19,340 --> 00:36:21,079 you will be able to use this back door, 1028 00:36:21,080 --> 00:36:22,999 that you'll have the key to keep others 1029 00:36:23,000 --> 00:36:25,339 out and have access for yourself. 1030 00:36:25,340 --> 00:36:27,469 And yet what we see here is an example 1031 00:36:27,470 --> 00:36:29,209 where someone was able to switch out the 1032 00:36:29,210 --> 00:36:31,339 locks, create a new key and 1033 00:36:31,340 --> 00:36:33,799 create their own back door that would 1034 00:36:33,800 --> 00:36:36,139 be dependent on this infrastructure 1035 00:36:36,140 --> 00:36:37,340 that had been maintained. 1036 00:36:39,360 --> 00:36:41,699 Finally, malware malware continues 1037 00:36:41,700 --> 00:36:43,799 to be a way 1038 00:36:43,800 --> 00:36:45,929 to root around encryption, 1039 00:36:45,930 --> 00:36:48,089 if you control the endpoint, you can 1040 00:36:48,090 --> 00:36:50,339 go ahead and look at the plain 1041 00:36:50,340 --> 00:36:50,619 text. 1042 00:36:50,620 --> 00:36:52,229 You use a key logger to get people's 1043 00:36:52,230 --> 00:36:54,659 passwords. You can basically 1044 00:36:54,660 --> 00:36:56,369 avoid the encryption so they can think 1045 00:36:56,370 --> 00:36:58,109 that they're having and end encryption. 1046 00:36:58,110 --> 00:37:00,599 But of course, only it's only good 1047 00:37:00,600 --> 00:37:02,730 if the end isn't compromised. 1048 00:37:04,670 --> 00:37:06,379 And this is this is a tool that is more 1049 00:37:06,380 --> 00:37:08,569 oriented towards targeted attacks, 1050 00:37:08,570 --> 00:37:10,699 while the other tools, like passively 1051 00:37:10,700 --> 00:37:13,039 looking at VPN, more oriented 1052 00:37:13,040 --> 00:37:14,359 towards mass attacks. 1053 00:37:15,380 --> 00:37:17,029 All right, so how do we fight back? 1054 00:37:17,030 --> 00:37:18,559 What are what are the arguments we can 1055 00:37:18,560 --> 00:37:20,989 raise? What can we say, the policymakers? 1056 00:37:20,990 --> 00:37:23,569 What can we do to fight against this? 1057 00:37:23,570 --> 00:37:25,369 Well, we can rely on arguments with 1058 00:37:25,370 --> 00:37:27,589 principle, with public policy, 1059 00:37:27,590 --> 00:37:30,859 with pragmatism and with promotion, 1060 00:37:30,860 --> 00:37:32,149 with principle. 1061 00:37:32,150 --> 00:37:33,619 We have to explain to people that strong 1062 00:37:33,620 --> 00:37:35,809 encryption is required to 1063 00:37:35,810 --> 00:37:37,879 effectuate human rights principles, that 1064 00:37:37,880 --> 00:37:39,439 we need it for privacy, that we need it 1065 00:37:39,440 --> 00:37:41,689 for free expression that encryption 1066 00:37:41,690 --> 00:37:43,549 is going to help us bring a brighter 1067 00:37:43,550 --> 00:37:45,649 future and look to the 1068 00:37:45,650 --> 00:37:47,689 Universal Declaration of Human Rights, 1069 00:37:47,690 --> 00:37:49,759 which enshrines within it the 1070 00:37:49,760 --> 00:37:51,439 notion that people should have the 1071 00:37:51,440 --> 00:37:53,629 privacy free from arbitrary 1072 00:37:53,630 --> 00:37:54,739 interference. 1073 00:37:55,760 --> 00:37:58,369 The principle from Article 19 1074 00:37:58,370 --> 00:38:00,349 free expression that people have the 1075 00:38:00,350 --> 00:38:01,999 right to free expression that right 1076 00:38:02,000 --> 00:38:04,159 includes the ability to seek, 1077 00:38:04,160 --> 00:38:06,259 receive and impart information and 1078 00:38:06,260 --> 00:38:09,079 ideas through any media. 1079 00:38:09,080 --> 00:38:11,239 And in order to effectuate that right 1080 00:38:11,240 --> 00:38:13,939 to receive and impart information 1081 00:38:13,940 --> 00:38:16,339 and ideas, we need to have encryption. 1082 00:38:16,340 --> 00:38:18,379 We need, like any media should include 1083 00:38:18,380 --> 00:38:20,539 encryption. And you should be able 1084 00:38:20,540 --> 00:38:23,239 to include within those ideas 1085 00:38:23,240 --> 00:38:24,169 encryption. 1086 00:38:24,170 --> 00:38:26,239 And this is also the notion that code is 1087 00:38:26,240 --> 00:38:28,369 speech. Code is an aspect of freedom of 1088 00:38:28,370 --> 00:38:30,169 expression, that if we are going to have 1089 00:38:30,170 --> 00:38:31,639 these rights as important human right 1090 00:38:31,640 --> 00:38:33,889 principle, then it must allow for 1091 00:38:33,890 --> 00:38:36,049 the publication of strong 1092 00:38:36,050 --> 00:38:38,929 and Christian crypto systems. 1093 00:38:38,930 --> 00:38:40,670 And I think this especially comes 1094 00:38:41,780 --> 00:38:44,149 true for open source systems 1095 00:38:44,150 --> 00:38:45,499 where people are putting out and 1096 00:38:45,500 --> 00:38:48,139 publishing things for the world to see 1097 00:38:48,140 --> 00:38:50,149 because they want the world to look at 1098 00:38:50,150 --> 00:38:51,829 their crypto systems, see what the 1099 00:38:51,830 --> 00:38:54,289 problems might be, find vulnerabilities, 1100 00:38:54,290 --> 00:38:55,279 report them back. 1101 00:38:55,280 --> 00:38:57,199 Those things can be improved and be part 1102 00:38:57,200 --> 00:38:58,819 of the ongoing dialog. 1103 00:38:58,820 --> 00:39:01,519 And this requires having 1104 00:39:01,520 --> 00:39:03,079 the freedom of publication, the freedom 1105 00:39:03,080 --> 00:39:04,909 of discussion, and having people come 1106 00:39:04,910 --> 00:39:06,949 together at places like C.C.C. 1107 00:39:06,950 --> 00:39:09,589 to discuss, debate and improve 1108 00:39:09,590 --> 00:39:11,499 on the crypto systems. 1109 00:39:11,500 --> 00:39:12,859 And I think the other important human 1110 00:39:12,860 --> 00:39:14,419 right principle that needs to be 1111 00:39:14,420 --> 00:39:16,699 recognized is that protecting against 1112 00:39:16,700 --> 00:39:19,489 oppressive regimes is more important 1113 00:39:19,490 --> 00:39:21,589 than maximizing spying, 1114 00:39:21,590 --> 00:39:23,989 that we have a greater role in trying 1115 00:39:23,990 --> 00:39:26,239 to protect people, give them the 1116 00:39:26,240 --> 00:39:28,279 freedom to organize, to talk among 1117 00:39:28,280 --> 00:39:30,409 themselves, to have a private space 1118 00:39:30,410 --> 00:39:32,509 to effectuate their 1119 00:39:32,510 --> 00:39:35,059 democracy rather than 1120 00:39:35,060 --> 00:39:37,489 increasing and maximizing the power 1121 00:39:37,490 --> 00:39:38,490 of the state. 1122 00:39:47,700 --> 00:39:50,399 So weakening encryption 1123 00:39:50,400 --> 00:39:52,559 is actually it's mostly good for 1124 00:39:52,560 --> 00:39:54,749 mass on targeted spying, 1125 00:39:54,750 --> 00:39:56,909 things like where it is, 1126 00:39:58,320 --> 00:39:59,489 if they have target is fine. 1127 00:39:59,490 --> 00:40:01,979 They can use tools like malware, targeted 1128 00:40:01,980 --> 00:40:04,079 attacks, tailored 1129 00:40:04,080 --> 00:40:06,149 access and go 1130 00:40:06,150 --> 00:40:07,919 after a particular target. 1131 00:40:07,920 --> 00:40:09,509 But when you're going after an entire 1132 00:40:09,510 --> 00:40:11,669 crypto system, what is the purpose 1133 00:40:11,670 --> 00:40:13,889 of that? And the purpose is 1134 00:40:13,890 --> 00:40:15,989 to enable mass decryption of 1135 00:40:15,990 --> 00:40:18,029 mass communications so you control 1136 00:40:18,030 --> 00:40:20,459 through it and look for subversive 1137 00:40:20,460 --> 00:40:21,989 elements, look for keywords, 1138 00:40:23,340 --> 00:40:25,919 and mass spying is less effective 1139 00:40:25,920 --> 00:40:27,359 and more invasive. 1140 00:40:27,360 --> 00:40:29,009 So one of the reasons we have to hold the 1141 00:40:29,010 --> 00:40:31,139 line and having strong encryption is 1142 00:40:31,140 --> 00:40:33,389 at least to require a security 1143 00:40:33,390 --> 00:40:35,309 state to have to have a reason to go 1144 00:40:35,310 --> 00:40:37,889 after people have to put into some effort 1145 00:40:37,890 --> 00:40:39,449 and make it sure that they're only 1146 00:40:39,450 --> 00:40:42,299 targeting where they have the strongest 1147 00:40:42,300 --> 00:40:44,669 its most necessary to the 1148 00:40:44,670 --> 00:40:45,900 to the reasons they're looking for. 1149 00:40:47,250 --> 00:40:49,379 We also want to have strong encryption 1150 00:40:49,380 --> 00:40:52,259 so that we can have a feeling of 1151 00:40:52,260 --> 00:40:53,849 the strength coming from looking at the 1152 00:40:53,850 --> 00:40:56,339 math, looking at 1153 00:40:56,340 --> 00:40:58,529 how the crypto system works, having that 1154 00:40:58,530 --> 00:41:01,079 tested by our best cryptanalysis 1155 00:41:01,080 --> 00:41:03,269 so that we can say it's not just a black 1156 00:41:03,270 --> 00:41:05,369 box, that we put something in their trust 1157 00:41:05,370 --> 00:41:08,339 that it works OK, and then 1158 00:41:08,340 --> 00:41:09,689 hope for the best. 1159 00:41:09,690 --> 00:41:11,939 And the thing is, you can't combine that 1160 00:41:11,940 --> 00:41:14,309 with with back doors, because 1161 00:41:14,310 --> 00:41:16,499 if you're if you're putting in a backdoor 1162 00:41:16,500 --> 00:41:18,659 system, you don't know 1163 00:41:18,660 --> 00:41:20,579 all the ways in which it works out. 1164 00:41:20,580 --> 00:41:22,679 If you have something like the Clipper 1165 00:41:22,680 --> 00:41:24,899 chip, it was it was a black 1166 00:41:24,900 --> 00:41:27,239 box for the first three 1167 00:41:27,240 --> 00:41:28,859 or four years that it was available. 1168 00:41:28,860 --> 00:41:30,629 They weren't showing the algorithm. 1169 00:41:30,630 --> 00:41:32,429 It was just trust us, we've done this. 1170 00:41:32,430 --> 00:41:33,599 Well said. 1171 00:41:33,600 --> 00:41:35,519 We want strong encryption, which can be 1172 00:41:35,520 --> 00:41:37,679 looked at, tested and understood so 1173 00:41:37,680 --> 00:41:39,809 we can trust the math. 1174 00:41:39,810 --> 00:41:42,119 Also, we want strong encryption because 1175 00:41:42,120 --> 00:41:43,739 it enables innovation. 1176 00:41:43,740 --> 00:41:46,049 Since we we had the first crypto 1177 00:41:46,050 --> 00:41:48,509 wars, the availability 1178 00:41:48,510 --> 00:41:50,369 of strong encryption has been 1179 00:41:50,370 --> 00:41:52,409 tremendously useful. 1180 00:41:52,410 --> 00:41:54,569 The e-commerce that we use day 1181 00:41:54,570 --> 00:41:57,059 to day it was enabled by the ability 1182 00:41:57,060 --> 00:41:58,949 to have encrypted communications, 1183 00:41:58,950 --> 00:42:01,259 authenticated communications between 1184 00:42:01,260 --> 00:42:03,479 servers and browsers. 1185 00:42:03,480 --> 00:42:05,459 Strong encryption and the availability of 1186 00:42:05,460 --> 00:42:07,229 it have come with innovations like 1187 00:42:07,230 --> 00:42:09,389 Bitcoin. The possibility of 1188 00:42:09,390 --> 00:42:11,909 a cryptocurrency 1189 00:42:11,910 --> 00:42:14,069 depends on the availability of strong 1190 00:42:14,070 --> 00:42:16,169 encryption. And after having that be 1191 00:42:16,170 --> 00:42:18,839 a theoretical possibility 1192 00:42:18,840 --> 00:42:21,419 for many years, we have found in Bitcoin 1193 00:42:21,420 --> 00:42:23,909 a real world example of some innovation 1194 00:42:23,910 --> 00:42:25,829 relying upon strong encryption. 1195 00:42:28,740 --> 00:42:30,869 And if we go ahead and try to force the 1196 00:42:30,870 --> 00:42:33,749 companies to compromise on security, 1197 00:42:33,750 --> 00:42:36,149 this makes everybody less 1198 00:42:36,150 --> 00:42:38,339 safe, that 1199 00:42:38,340 --> 00:42:40,439 encryption is critical for 1200 00:42:40,440 --> 00:42:42,869 security. So oftentimes the policymakers, 1201 00:42:42,870 --> 00:42:45,179 when they're trying to argue encryption 1202 00:42:45,180 --> 00:42:47,339 against encryption, they're making an 1203 00:42:47,340 --> 00:42:49,409 appeal to national security. 1204 00:42:49,410 --> 00:42:51,779 We need to have weak encryption for 1205 00:42:51,780 --> 00:42:52,889 security. 1206 00:42:52,890 --> 00:42:54,269 This is a false debate. 1207 00:42:54,270 --> 00:42:57,239 This is there's not a trade off between 1208 00:42:57,240 --> 00:42:58,829 security and encryption. 1209 00:42:58,830 --> 00:43:00,989 Security is vital. 1210 00:43:00,990 --> 00:43:03,090 Encryption is vital for security. 1211 00:43:04,350 --> 00:43:06,419 At the same time, if you say, 1212 00:43:06,420 --> 00:43:08,190 all right, you know, we we 1213 00:43:09,270 --> 00:43:10,709 we still we want to have 1214 00:43:12,210 --> 00:43:13,199 this backdoor. 1215 00:43:13,200 --> 00:43:15,129 If you put aside the arguments about it, 1216 00:43:15,130 --> 00:43:17,609 even if you think that a backdoor 1217 00:43:17,610 --> 00:43:19,739 is a good idea, it's still 1218 00:43:19,740 --> 00:43:21,269 you have to address the other question. 1219 00:43:21,270 --> 00:43:23,819 Who do you give the back door to? 1220 00:43:23,820 --> 00:43:25,679 So even if you think your own government 1221 00:43:25,680 --> 00:43:27,989 is the best, most honorable government 1222 00:43:27,990 --> 00:43:30,269 in the world, and they will surely 1223 00:43:30,270 --> 00:43:32,549 only use this for for 1224 00:43:32,550 --> 00:43:34,110 happiness and good 1225 00:43:35,310 --> 00:43:36,929 at the same time. Well, what about the 1226 00:43:36,930 --> 00:43:38,729 other governments? And when you say that 1227 00:43:38,730 --> 00:43:41,039 we get to have a back door 1228 00:43:41,040 --> 00:43:43,259 for for these communications, 1229 00:43:43,260 --> 00:43:44,969 then the other governments, the ones that 1230 00:43:44,970 --> 00:43:46,409 you might not like, the ones you might 1231 00:43:46,410 --> 00:43:48,599 find to be a repressive regime, 1232 00:43:48,600 --> 00:43:51,509 they're going to ask for a backdoor, too. 1233 00:43:51,510 --> 00:43:53,909 And then who gets to have a how do we 1234 00:43:53,910 --> 00:43:56,219 have a principled discussion about 1235 00:43:56,220 --> 00:43:58,289 that? And we saw this, as we just saying, 1236 00:43:58,290 --> 00:44:00,389 in the Chinese anti-terrorism law, 1237 00:44:00,390 --> 00:44:03,359 where they're using citing to 1238 00:44:03,360 --> 00:44:05,879 some of the language and rhetoric 1239 00:44:05,880 --> 00:44:07,499 to say, well, we're just doing what other 1240 00:44:07,500 --> 00:44:09,299 countries are doing. 1241 00:44:09,300 --> 00:44:11,369 I think our public policy thing, you're 1242 00:44:11,370 --> 00:44:13,859 going dark is a big thing that that 1243 00:44:13,860 --> 00:44:15,539 you'll you'll hear from from governments 1244 00:44:15,540 --> 00:44:17,969 say, well, encryption is making us going 1245 00:44:17,970 --> 00:44:20,129 dark. And this this is just not 1246 00:44:20,130 --> 00:44:22,379 recognizing that we're actually in 1247 00:44:22,380 --> 00:44:24,449 a golden age of surveillance, 1248 00:44:24,450 --> 00:44:26,759 that right now you have cameras 1249 00:44:26,760 --> 00:44:29,159 on so many street corners, 1250 00:44:29,160 --> 00:44:31,169 people are carrying around a location 1251 00:44:31,170 --> 00:44:33,269 tracker, a mobile phone in their 1252 00:44:33,270 --> 00:44:35,039 pockets at all times. 1253 00:44:35,040 --> 00:44:37,109 Everything they do, making credit 1254 00:44:37,110 --> 00:44:39,449 card purchases, having 1255 00:44:39,450 --> 00:44:41,249 even encrypted communications or 1256 00:44:41,250 --> 00:44:43,449 providing tons of metadata that is being 1257 00:44:43,450 --> 00:44:45,509 trolled and analyzed, it 1258 00:44:45,510 --> 00:44:47,909 is is easy for governments 1259 00:44:47,910 --> 00:44:49,709 to conduct surveillance that has ever 1260 00:44:49,710 --> 00:44:51,959 been. And yet they want to 1261 00:44:51,960 --> 00:44:53,339 say that they're going dark. 1262 00:44:53,340 --> 00:44:55,439 Just add this extra element so they 1263 00:44:55,440 --> 00:44:57,029 can try and get the encrypted 1264 00:44:57,030 --> 00:44:58,030 communications. 1265 00:44:59,020 --> 00:45:00,779 Could also argue about pragmatism. 1266 00:45:00,780 --> 00:45:02,999 Well, I mean, one argument, it won't 1267 00:45:03,000 --> 00:45:04,049 work. 1268 00:45:04,050 --> 00:45:06,269 And this is true as far as it goes, 1269 00:45:06,270 --> 00:45:08,549 that when you if you if you try to ban 1270 00:45:08,550 --> 00:45:09,959 encryption, you're trying to say that you 1271 00:45:09,960 --> 00:45:12,269 can't have strong encryption. 1272 00:45:12,270 --> 00:45:13,499 What are you going to do about an open 1273 00:45:13,500 --> 00:45:15,659 source project? If you mandate that it 1274 00:45:15,660 --> 00:45:17,849 has a back door, someone might compile 1275 00:45:17,850 --> 00:45:19,199 it without that backdoor 1276 00:45:20,250 --> 00:45:22,409 and free software is hard 1277 00:45:22,410 --> 00:45:24,089 to stop. And even if you make it hard to 1278 00:45:24,090 --> 00:45:26,549 publish in certain jurisdictions, 1279 00:45:26,550 --> 00:45:28,269 information wants to be free. 1280 00:45:28,270 --> 00:45:30,420 It will find a way out there. 1281 00:45:31,590 --> 00:45:33,029 This is good as far as it goes. 1282 00:45:33,030 --> 00:45:35,609 Those we're saying before, 1283 00:45:35,610 --> 00:45:36,610 the governments 1284 00:45:37,710 --> 00:45:40,199 are mostly concerned with widespread 1285 00:45:40,200 --> 00:45:42,389 availability of encryption so 1286 00:45:42,390 --> 00:45:44,939 that it's only so good to have 1287 00:45:44,940 --> 00:45:46,439 encryption available with people, with 1288 00:45:46,440 --> 00:45:48,659 the technical sophistication enough 1289 00:45:48,660 --> 00:45:50,340 to compile their own code. 1290 00:45:51,480 --> 00:45:53,609 We could also argue on a pragmatic level 1291 00:45:53,610 --> 00:45:56,309 about math that it is simply not possible 1292 00:45:56,310 --> 00:45:58,619 to make encryption simultaneously 1293 00:45:58,620 --> 00:46:01,079 weak and strong. 1294 00:46:01,080 --> 00:46:02,639 And finally, if the argument is about 1295 00:46:02,640 --> 00:46:04,919 national security, about terrorists, all 1296 00:46:04,920 --> 00:46:07,979 these efforts to weaken encryption 1297 00:46:07,980 --> 00:46:10,499 will mostly affect law abiding people. 1298 00:46:10,500 --> 00:46:12,329 It's not going to stop terrorists from 1299 00:46:12,330 --> 00:46:14,070 being able to use these tools. 1300 00:46:16,280 --> 00:46:17,779 So what can you do about it? 1301 00:46:17,780 --> 00:46:19,699 Well, you can help by promoting, 1302 00:46:19,700 --> 00:46:22,219 creating, improving and using encryption, 1303 00:46:23,450 --> 00:46:26,449 show your friends how to use encryption. 1304 00:46:26,450 --> 00:46:27,439 Thank you. 1305 00:46:27,440 --> 00:46:28,440 Go to. 1306 00:46:32,750 --> 00:46:34,879 Make it as widespread as possible. 1307 00:46:34,880 --> 00:46:36,859 You know how to use it, I think probably 1308 00:46:36,860 --> 00:46:38,479 most people in this room have used 1309 00:46:38,480 --> 00:46:40,489 encryption many, many times. 1310 00:46:40,490 --> 00:46:41,869 But you might have some friends who are 1311 00:46:41,870 --> 00:46:43,699 new to this, show them how to use it to 1312 00:46:43,700 --> 00:46:45,889 show them how to operate a secure 1313 00:46:45,890 --> 00:46:48,079 messenger, show them how to 1314 00:46:48,080 --> 00:46:49,189 install TOR 1315 00:46:50,450 --> 00:46:53,329 has a program surveillance self-defense, 1316 00:46:53,330 --> 00:46:55,879 SD EFG. 1317 00:46:55,880 --> 00:46:58,009 It has tools to show people how to defend 1318 00:46:58,010 --> 00:46:59,539 themselves. 1319 00:46:59,540 --> 00:47:01,489 And for those of you who are programmers 1320 00:47:01,490 --> 00:47:02,689 who are working on some of these 1321 00:47:02,690 --> 00:47:04,969 projects, trying to make censorship 1322 00:47:04,970 --> 00:47:07,069 resistant crypto tools, make 1323 00:47:07,070 --> 00:47:08,510 them open source, make the 1324 00:47:10,130 --> 00:47:12,499 distribution as wide as possible 1325 00:47:12,500 --> 00:47:14,899 so that it will be difficult to take back 1326 00:47:14,900 --> 00:47:16,819 and put that genie back in the bottle, 1327 00:47:16,820 --> 00:47:19,549 use reproducible build so that someone 1328 00:47:19,550 --> 00:47:21,799 who download it will know they 1329 00:47:21,800 --> 00:47:23,899 are getting and compiling the 1330 00:47:23,900 --> 00:47:25,309 thing that you have shown them with the 1331 00:47:25,310 --> 00:47:27,530 open source to have strong encryption. 1332 00:47:28,970 --> 00:47:31,009 Just talk briefly about some of the 1333 00:47:31,010 --> 00:47:33,139 efforts that effort is working on to 1334 00:47:33,140 --> 00:47:35,509 try and help with the widespread adoption 1335 00:47:35,510 --> 00:47:36,799 of encryption. 1336 00:47:36,800 --> 00:47:38,629 You heard earlier in C.C.C.. 1337 00:47:38,630 --> 00:47:39,619 I hope some of you went to it. 1338 00:47:39,620 --> 00:47:41,569 There was a talk about let's encrypt. 1339 00:47:41,570 --> 00:47:43,729 This is a new certificate authority 1340 00:47:43,730 --> 00:47:46,249 to try to make it easy and fun 1341 00:47:46,250 --> 00:47:48,409 to add SSL to every 1342 00:47:48,410 --> 00:47:50,059 website, to add a transport layer of 1343 00:47:50,060 --> 00:47:52,999 security so that there's no more excuse 1344 00:47:53,000 --> 00:47:55,219 for having an unencrypted 1345 00:47:55,220 --> 00:47:56,220 website. 1346 00:47:59,950 --> 00:48:02,229 We also have been looking at 1347 00:48:02,230 --> 00:48:04,989 rating big providers 1348 00:48:04,990 --> 00:48:07,179 on how well they are working 1349 00:48:07,180 --> 00:48:09,279 at encrypting the Web, so we have the 1350 00:48:09,280 --> 00:48:11,079 encrypt the Web report, there's a 1351 00:48:11,080 --> 00:48:12,249 screenshot of it there. 1352 00:48:12,250 --> 00:48:15,369 You can see a number of companies have 1353 00:48:15,370 --> 00:48:17,139 gotten five out of five. 1354 00:48:17,140 --> 00:48:19,209 This is in part a reaction to that 1355 00:48:19,210 --> 00:48:21,999 smiley face we saw on the slide earlier 1356 00:48:22,000 --> 00:48:23,679 where they were putting in stronger 1357 00:48:23,680 --> 00:48:25,269 encryption in reaction to the student 1358 00:48:25,270 --> 00:48:27,099 document. And we're continuing to put 1359 00:48:27,100 --> 00:48:29,199 pressure on all the large providers 1360 00:48:29,200 --> 00:48:31,449 to massively increase their use of strong 1361 00:48:31,450 --> 00:48:32,450 encryption. 1362 00:48:33,400 --> 00:48:35,799 We also have the secure messaging 1363 00:48:35,800 --> 00:48:38,079 scorecard, which goes 1364 00:48:38,080 --> 00:48:40,239 through I think it's about 30 or 40 1365 00:48:40,240 --> 00:48:42,369 different messaging tools and 1366 00:48:42,370 --> 00:48:44,199 provides some information about what kind 1367 00:48:44,200 --> 00:48:46,419 of encryption they use with 1368 00:48:46,420 --> 00:48:47,689 a checkbox system. 1369 00:48:47,690 --> 00:48:49,929 So if you want to find out what messaging 1370 00:48:49,930 --> 00:48:51,549 systems are using encryption, how much 1371 00:48:51,550 --> 00:48:53,109 they're using, go check it out of the 1372 00:48:53,110 --> 00:48:55,329 secure messaging scorecard and 1373 00:48:55,330 --> 00:48:57,549 try to add and start using 1374 00:48:57,550 --> 00:48:59,859 as many encryption programs as possible 1375 00:48:59,860 --> 00:49:02,229 so that you can take advantage 1376 00:49:02,230 --> 00:49:04,359 of strong encryption, keep it strong, 1377 00:49:04,360 --> 00:49:07,149 keep it safe, and make sure that we win 1378 00:49:07,150 --> 00:49:08,949 crypto wars. Part two. 1379 00:49:08,950 --> 00:49:09,950 Thank you very much. 1380 00:49:21,890 --> 00:49:23,959 Thank you. Thank you for a great talk. 1381 00:49:23,960 --> 00:49:26,389 We now have about 10 minutes for Q&A, 1382 00:49:26,390 --> 00:49:28,039 so please just line up at the 1383 00:49:28,040 --> 00:49:29,249 microphones. 1384 00:49:29,250 --> 00:49:31,789 Of course, if you're on the Internet, 1385 00:49:31,790 --> 00:49:33,859 use ISEA or Twitter to 1386 00:49:33,860 --> 00:49:36,109 also ask questions. 1387 00:49:36,110 --> 00:49:38,419 And we will just. And if you're leaving, 1388 00:49:38,420 --> 00:49:40,249 please be quiet. 1389 00:49:40,250 --> 00:49:42,079 Stay out of the line of sight of the 1390 00:49:42,080 --> 00:49:43,399 cameras. 1391 00:49:43,400 --> 00:49:44,400 Just try to be. 1392 00:49:45,750 --> 00:49:47,879 Just don't leave maybe, but be quiet if 1393 00:49:47,880 --> 00:49:48,839 you do. 1394 00:49:48,840 --> 00:49:50,610 I now microphone number one, please. 1395 00:49:52,350 --> 00:49:55,049 I I happened to 1396 00:49:55,050 --> 00:49:57,539 stumble upon the scorecard 1397 00:49:57,540 --> 00:49:59,789 once and 1398 00:49:59,790 --> 00:50:01,949 and WhatsApp was marked 1399 00:50:01,950 --> 00:50:05,010 with a lot of green checkboxes, 1400 00:50:06,910 --> 00:50:09,479 how much can we trust WhatsApp, 1401 00:50:09,480 --> 00:50:11,549 especially in the terms of it 1402 00:50:11,550 --> 00:50:13,649 being a really high 1403 00:50:13,650 --> 00:50:16,469 value target since 1404 00:50:16,470 --> 00:50:17,830 so many people are using it? 1405 00:50:18,990 --> 00:50:20,879 So what's out there is indeed a high 1406 00:50:20,880 --> 00:50:23,219 value target? I think they have, 1407 00:50:23,220 --> 00:50:25,169 because they're saying over a billion 1408 00:50:25,170 --> 00:50:27,599 users for a long time they 1409 00:50:27,600 --> 00:50:30,299 were not having strong encryption. 1410 00:50:30,300 --> 00:50:32,549 They started to work with 1411 00:50:32,550 --> 00:50:34,709 open whisper systems and Moxy 1412 00:50:34,710 --> 00:50:36,809 Marlinspike to increase 1413 00:50:36,810 --> 00:50:38,969 the encryption right around the time that 1414 00:50:38,970 --> 00:50:41,219 they were being absorbed into 1415 00:50:41,220 --> 00:50:42,220 Facebook. 1416 00:50:43,350 --> 00:50:45,299 So I think they're they are making 1417 00:50:45,300 --> 00:50:47,669 efforts to increase the encryption. 1418 00:50:47,670 --> 00:50:50,099 Have it have it be good, 1419 00:50:50,100 --> 00:50:52,499 though? I would say the secure messaging 1420 00:50:52,500 --> 00:50:53,500 scorecard. 1421 00:50:54,810 --> 00:50:56,909 I think this is true actually of a number 1422 00:50:56,910 --> 00:50:59,429 of our rating things. It it does rely 1423 00:50:59,430 --> 00:51:01,589 upon public information 1424 00:51:01,590 --> 00:51:03,059 that's available. And we don't have the 1425 00:51:03,060 --> 00:51:05,159 ability to to sort of go in there 1426 00:51:05,160 --> 00:51:07,409 and see if something secret has happened 1427 00:51:07,410 --> 00:51:09,539 to look through the to unless there's 1428 00:51:09,540 --> 00:51:11,249 open source code, we can't look through 1429 00:51:11,250 --> 00:51:12,179 the code. 1430 00:51:12,180 --> 00:51:14,309 So it's based upon the information 1431 00:51:14,310 --> 00:51:15,929 that is available. 1432 00:51:15,930 --> 00:51:17,669 So I am hopeful that that they are 1433 00:51:17,670 --> 00:51:19,140 looking at that, but 1434 00:51:20,190 --> 00:51:22,669 only they can can guarantee. 1435 00:51:22,670 --> 00:51:24,160 So sorry. 1436 00:51:25,350 --> 00:51:27,899 So you did not talk 1437 00:51:27,900 --> 00:51:30,149 to the internal auditors that 1438 00:51:30,150 --> 00:51:31,709 happened to have a look at the source 1439 00:51:31,710 --> 00:51:32,609 code? 1440 00:51:32,610 --> 00:51:33,899 That's correct. OK, thank you. 1441 00:51:35,220 --> 00:51:36,779 And now a question from the Internet, 1442 00:51:36,780 --> 00:51:38,859 please, thank you. 1443 00:51:38,860 --> 00:51:40,709 Are we facing a future where encryption 1444 00:51:40,710 --> 00:51:42,629 is totally prohibited by law? 1445 00:51:42,630 --> 00:51:44,849 And why is it not 1446 00:51:44,850 --> 00:51:46,949 today? So is it because the agency 1447 00:51:46,950 --> 00:51:48,859 are still able to obtain the data? 1448 00:51:51,500 --> 00:51:53,599 So are we facing a 1449 00:51:53,600 --> 00:51:55,159 future in which encryption will be here, 1450 00:51:55,160 --> 00:51:57,289 but no, not yet 1451 00:51:57,290 --> 00:51:59,359 and I hope not ever. 1452 00:51:59,360 --> 00:52:01,939 We won the first crypto wars 1453 00:52:01,940 --> 00:52:03,739 and I think we can win this one. 1454 00:52:03,740 --> 00:52:05,719 We can show them both through, through 1455 00:52:05,720 --> 00:52:07,819 reasoned, through, through principle, 1456 00:52:07,820 --> 00:52:08,809 through rhetoric. 1457 00:52:08,810 --> 00:52:10,729 Why banning encryption? 1458 00:52:10,730 --> 00:52:12,259 Having a world without a strong 1459 00:52:12,260 --> 00:52:14,689 encryption is a terrible idea. 1460 00:52:14,690 --> 00:52:16,759 And we will fight to maintain 1461 00:52:16,760 --> 00:52:18,199 that as long as we can. 1462 00:52:18,200 --> 00:52:20,329 And even if some jurisdictions 1463 00:52:20,330 --> 00:52:22,699 do pass laws that ban 1464 00:52:22,700 --> 00:52:24,799 encryption, encryption will still be out 1465 00:52:24,800 --> 00:52:27,409 there, be available from other sites. 1466 00:52:27,410 --> 00:52:29,119 So I think that we can win this war. 1467 00:52:36,100 --> 00:52:37,999 Microphone number two, please. 1468 00:52:38,000 --> 00:52:40,089 Oh, hi, my name is Mary, 1469 00:52:40,090 --> 00:52:42,339 and this should not be intended to be 1470 00:52:42,340 --> 00:52:45,129 the troll question, but what the other 1471 00:52:45,130 --> 00:52:47,169 one is opinion about the homebrewed 1472 00:52:47,170 --> 00:52:48,170 krypto. 1473 00:52:48,820 --> 00:52:51,939 So my opinion about Homebrew Krypto. 1474 00:52:51,940 --> 00:52:54,009 Well, one thing I think 1475 00:52:54,010 --> 00:52:56,169 Bruce Schneier has said is 1476 00:52:56,170 --> 00:52:58,599 that anybody can create a crypto system 1477 00:52:58,600 --> 00:53:00,850 that they themselves cannot break 1478 00:53:03,100 --> 00:53:04,100 and. 1479 00:53:08,240 --> 00:53:10,009 So, I mean, this is not to say that 1480 00:53:10,010 --> 00:53:12,319 someone can't come up with a 1481 00:53:12,320 --> 00:53:14,809 good idea, but before 1482 00:53:14,810 --> 00:53:16,969 you can put much trust 1483 00:53:16,970 --> 00:53:19,039 into these things, it has to be peer 1484 00:53:19,040 --> 00:53:21,439 reviewed. It has to be made available, 1485 00:53:21,440 --> 00:53:24,139 get world class cryptographers 1486 00:53:24,140 --> 00:53:26,869 to attack it, find the flaws, 1487 00:53:26,870 --> 00:53:30,109 improve it based upon those flaws, 1488 00:53:30,110 --> 00:53:32,479 and see if you can do better. 1489 00:53:32,480 --> 00:53:35,149 So, you know, if somebody has 1490 00:53:35,150 --> 00:53:38,119 a better a better crypto system, 1491 00:53:38,120 --> 00:53:40,219 the only way we're going to know is 1492 00:53:40,220 --> 00:53:42,559 if it is heavily tested, heavily 1493 00:53:42,560 --> 00:53:44,689 vetted and not put out there 1494 00:53:44,690 --> 00:53:46,969 until everybody can look at it, 1495 00:53:46,970 --> 00:53:49,639 attack it and fail to break it. 1496 00:53:49,640 --> 00:53:51,869 But what if the crypto is using just only 1497 00:53:51,870 --> 00:53:54,589 those private circle of people? 1498 00:53:54,590 --> 00:53:56,849 If the crypto is only sorry 1499 00:53:56,850 --> 00:53:59,239 for some some for for a small group 1500 00:53:59,240 --> 00:54:01,339 of people, I would 1501 00:54:01,340 --> 00:54:03,589 still rely upon publicly 1502 00:54:03,590 --> 00:54:05,449 vetted crypto programs. 1503 00:54:05,450 --> 00:54:07,369 And if you're if you're creating an 1504 00:54:07,370 --> 00:54:09,529 application that uses crypto, 1505 00:54:09,530 --> 00:54:12,529 you should not only use a 1506 00:54:12,530 --> 00:54:14,179 crypto program that has been tested, 1507 00:54:14,180 --> 00:54:16,579 explained why you're choosing that one, 1508 00:54:16,580 --> 00:54:18,919 have some some thought 1509 00:54:18,920 --> 00:54:20,359 into it so people can understand that 1510 00:54:20,360 --> 00:54:22,429 decision making process and make sure 1511 00:54:22,430 --> 00:54:23,749 they're making their decision about 1512 00:54:23,750 --> 00:54:26,309 whether to use it based upon it. 1513 00:54:26,310 --> 00:54:27,310 OK, great, thanks. 1514 00:54:31,180 --> 00:54:33,399 Number five, please. 1515 00:54:33,400 --> 00:54:35,199 Yeah, hi, my name is Mathias, I'm on the 1516 00:54:35,200 --> 00:54:36,549 board of Reporters Without Borders, 1517 00:54:36,550 --> 00:54:38,649 Germany, and in case you don't 1518 00:54:38,650 --> 00:54:40,929 know, here we are suing the BND because 1519 00:54:40,930 --> 00:54:42,999 we think that what they're doing in 1520 00:54:43,000 --> 00:54:45,549 collecting data and analyzing 1521 00:54:45,550 --> 00:54:48,009 them is out of proportion, 1522 00:54:48,010 --> 00:54:49,539 breaking the law and probably even the 1523 00:54:49,540 --> 00:54:50,499 Constitution. 1524 00:54:50,500 --> 00:54:51,959 So thank you for that. 1525 00:54:51,960 --> 00:54:52,960 Yeah. 1526 00:54:58,010 --> 00:55:00,139 Now, I gave that little intro because 1527 00:55:00,140 --> 00:55:01,609 I have a question, it might not be so 1528 00:55:01,610 --> 00:55:03,709 popular here in this room, 1529 00:55:03,710 --> 00:55:06,079 if you assume that your conclusions 1530 00:55:06,080 --> 00:55:07,969 and your propositions are correct and I 1531 00:55:07,970 --> 00:55:10,339 think they are, then we 1532 00:55:10,340 --> 00:55:12,169 have to assume that we need strong 1533 00:55:12,170 --> 00:55:14,779 crypto. If we at the same time assume 1534 00:55:14,780 --> 00:55:16,579 end to end encryption, if we at the same 1535 00:55:16,580 --> 00:55:18,739 time assume that law 1536 00:55:18,740 --> 00:55:20,899 enforcement and police have to have some 1537 00:55:20,900 --> 00:55:23,299 way to go after the bad guys, 1538 00:55:23,300 --> 00:55:24,409 what's the consequence? 1539 00:55:24,410 --> 00:55:26,539 Does that mean that you are 1540 00:55:26,540 --> 00:55:28,879 in favor of giving law enforcement 1541 00:55:28,880 --> 00:55:31,609 the ability, for example, to use 1542 00:55:31,610 --> 00:55:33,349 Trojan horses or other kinds of 1543 00:55:33,350 --> 00:55:35,209 technologies to put them on people's 1544 00:55:35,210 --> 00:55:37,459 computers? Because we can assume 1545 00:55:37,460 --> 00:55:39,829 that the bad guys are not just doing 1546 00:55:39,830 --> 00:55:42,109 phone calls over regular 1547 00:55:42,110 --> 00:55:44,209 phone lines nowadays, or if you don't 1548 00:55:44,210 --> 00:55:47,089 think that this is proportionate 1549 00:55:47,090 --> 00:55:49,899 or the way to go, then what is. 1550 00:55:49,900 --> 00:55:52,249 Well, thank you for that question. 1551 00:55:54,400 --> 00:55:56,709 But what we've done at EFF 1552 00:55:56,710 --> 00:55:58,249 with ourselves, actually a number of 1553 00:55:58,250 --> 00:56:00,529 other organizations that we can come up 1554 00:56:00,530 --> 00:56:02,869 with the necessary and proportionate 1555 00:56:02,870 --> 00:56:04,759 principles you go to necessary and 1556 00:56:04,760 --> 00:56:06,289 proportionate dawg to see them. 1557 00:56:06,290 --> 00:56:08,809 There are 13 principles for 1558 00:56:08,810 --> 00:56:10,789 government surveillance in order to 1559 00:56:10,790 --> 00:56:13,309 balance the needs of the state 1560 00:56:13,310 --> 00:56:15,419 against human rights principles. 1561 00:56:15,420 --> 00:56:17,719 So when they can go forward, I mean, 1562 00:56:17,720 --> 00:56:19,639 there are 13 principles. 1563 00:56:19,640 --> 00:56:21,469 It's a bit much to go through right now. 1564 00:56:21,470 --> 00:56:23,689 But they the gist of it is that 1565 00:56:23,690 --> 00:56:25,429 we want to make sure that it's only when 1566 00:56:25,430 --> 00:56:26,539 it is necessary. 1567 00:56:26,540 --> 00:56:28,939 And the amount 1568 00:56:28,940 --> 00:56:30,829 that the government will be allowed to do 1569 00:56:30,830 --> 00:56:32,959 is proportionate to the crime that 1570 00:56:32,960 --> 00:56:34,919 they're investigating the the act that 1571 00:56:34,920 --> 00:56:37,039 they're they're investigating so 1572 00:56:37,040 --> 00:56:38,629 that whether they may use a particular 1573 00:56:38,630 --> 00:56:40,069 tool would be dependent upon the 1574 00:56:40,070 --> 00:56:42,709 circumstances and always with 1575 00:56:42,710 --> 00:56:44,839 court oversight, making sure 1576 00:56:44,840 --> 00:56:47,209 that it is done in accordance with law, 1577 00:56:47,210 --> 00:56:49,219 in accordance with these principles. 1578 00:56:49,220 --> 00:56:50,929 So I encourage you to check out necessary 1579 00:56:50,930 --> 00:56:52,489 and proportionate, dawg. 1580 00:56:52,490 --> 00:56:53,490 Thanks. 1581 00:56:57,070 --> 00:57:00,429 And and now the Internet's place. 1582 00:57:00,430 --> 00:57:02,679 Are there any documents, ATF, 1583 00:57:02,680 --> 00:57:04,839 about CUFI and Jakob talked about 1584 00:57:04,840 --> 00:57:07,179 a while ago about Q 1585 00:57:07,180 --> 00:57:09,519 Fire validation software? 1586 00:57:09,520 --> 00:57:10,520 I'm not aware of any. 1587 00:57:12,380 --> 00:57:14,189 Well, that was quick number four, please. 1588 00:57:15,530 --> 00:57:17,659 I really like the overview of arguments 1589 00:57:17,660 --> 00:57:20,149 you you gave. I discuss these matters 1590 00:57:20,150 --> 00:57:21,799 often with a wide range of people in the 1591 00:57:21,800 --> 00:57:23,299 past few months. 1592 00:57:23,300 --> 00:57:25,159 And one of the arguments that you have 1593 00:57:25,160 --> 00:57:27,289 not actively or directly refute is 1594 00:57:27,290 --> 00:57:29,539 that, well, we have wiretapping 1595 00:57:29,540 --> 00:57:32,309 laws for phone conversations. 1596 00:57:32,310 --> 00:57:34,399 And could they 1597 00:57:34,400 --> 00:57:36,559 not just analogously holds 1598 00:57:36,560 --> 00:57:38,209 for digital communications in a more 1599 00:57:38,210 --> 00:57:40,159 broad sense, what we would be your 1600 00:57:40,160 --> 00:57:42,189 response to such an argument? 1601 00:57:42,190 --> 00:57:43,609 But I think there's actually less and 1602 00:57:43,610 --> 00:57:45,229 less of a difference between phone 1603 00:57:45,230 --> 00:57:47,239 conversations and electronic 1604 00:57:47,240 --> 00:57:49,039 communications. In fact, these days, many 1605 00:57:49,040 --> 00:57:50,749 phone conversations are actually going 1606 00:57:50,750 --> 00:57:52,999 over voice over IP there. 1607 00:57:53,000 --> 00:57:54,649 They are being transmitted the same way 1608 00:57:54,650 --> 00:57:56,809 as as an electronic 1609 00:57:56,810 --> 00:57:58,639 communication. 1610 00:57:58,640 --> 00:58:00,739 And so, I mean, and 1611 00:58:00,740 --> 00:58:02,209 many countries actually have brought 1612 00:58:02,210 --> 00:58:04,339 these things together and applied some 1613 00:58:04,340 --> 00:58:06,499 of the wiretapping rules to 1614 00:58:06,500 --> 00:58:08,899 voice over IP, even though it 1615 00:58:08,900 --> 00:58:11,859 goes over a different, different network. 1616 00:58:11,860 --> 00:58:12,919 You know, the important thing, the 1617 00:58:12,920 --> 00:58:15,199 principles behind this is to make sure 1618 00:58:15,200 --> 00:58:17,389 that they are getting access 1619 00:58:17,390 --> 00:58:20,149 to the content of communications 1620 00:58:20,150 --> 00:58:22,069 only in accordance to law with a high 1621 00:58:22,070 --> 00:58:23,569 standard having, you know, come back with 1622 00:58:23,570 --> 00:58:25,819 a warrant, make sure that before 1623 00:58:25,820 --> 00:58:28,279 you get access to to voice communications 1624 00:58:28,280 --> 00:58:30,409 or written communications, you 1625 00:58:30,410 --> 00:58:32,179 are meeting the highest possible test to 1626 00:58:32,180 --> 00:58:34,279 make sure that this is something that is 1627 00:58:34,280 --> 00:58:36,739 necessary and proportionate to the 1628 00:58:36,740 --> 00:58:38,189 investigation. 1629 00:58:38,190 --> 00:58:40,249 So the reason I 1630 00:58:40,250 --> 00:58:42,079 bring this up is because often it's not 1631 00:58:42,080 --> 00:58:44,279 allowed to construct 1632 00:58:44,280 --> 00:58:46,039 a telecommunications network in such a 1633 00:58:46,040 --> 00:58:48,349 way that it's not possible to wiretap 1634 00:58:48,350 --> 00:58:50,059 the plane conversation going on. 1635 00:58:50,060 --> 00:58:52,249 And this is in some sense analogous to 1636 00:58:52,250 --> 00:58:53,779 what you're discussing here. 1637 00:58:53,780 --> 00:58:55,429 It is. I mean, I think, you know, in the 1638 00:58:55,430 --> 00:58:56,989 United States, we have the computer 1639 00:58:56,990 --> 00:58:58,669 assistance, the Law Enforcement Act, 1640 00:58:58,670 --> 00:59:00,979 KHALIA, it was actually passed in 1641 00:59:00,980 --> 00:59:02,329 the 90s. 1642 00:59:02,330 --> 00:59:04,579 It exempted the Internet, but for 1643 00:59:04,580 --> 00:59:06,979 voice communications required 1644 00:59:06,980 --> 00:59:09,559 to have some some access 1645 00:59:09,560 --> 00:59:12,409 to these to these communications. 1646 00:59:12,410 --> 00:59:15,199 And so I don't think we should 1647 00:59:15,200 --> 00:59:16,339 move that beyond that. 1648 00:59:16,340 --> 00:59:17,719 I thought we actually and I think there 1649 00:59:17,720 --> 00:59:19,159 was a terrible idea in the beginning, 1650 00:59:21,260 --> 00:59:23,449 and this still does not 1651 00:59:23,450 --> 00:59:24,589 prevent you from using. 1652 00:59:24,590 --> 00:59:26,119 And encryption is just the people don't 1653 00:59:26,120 --> 00:59:28,459 really have 1654 00:59:28,460 --> 00:59:30,679 easy technologies if you're using a plain 1655 00:59:30,680 --> 00:59:32,809 old telephone system to have an encrypted 1656 00:59:32,810 --> 00:59:33,859 communication. 1657 00:59:33,860 --> 00:59:36,349 But if you use a 1658 00:59:36,350 --> 00:59:38,449 voice over IP application, for example, 1659 00:59:38,450 --> 00:59:40,729 if you use signal, you can have 1660 00:59:40,730 --> 00:59:43,159 an encrypted voice communication 1661 00:59:43,160 --> 00:59:45,020 so that even if they have a 1662 00:59:46,220 --> 00:59:48,649 wiretap compliant network, 1663 00:59:48,650 --> 00:59:49,789 they're still just going to get the 1664 00:59:49,790 --> 00:59:51,919 encrypted information that was 1665 00:59:51,920 --> 00:59:54,079 going through. So end and encryption 1666 00:59:54,080 --> 00:59:56,149 is the is the better solution for 1667 00:59:56,150 --> 00:59:57,150 the end user. 1668 00:59:58,570 --> 00:59:59,829 And we're out of time. 1669 00:59:59,830 --> 01:00:01,899 Please think of it again. 1670 01:00:01,900 --> 01:00:02,900 Thank you very much.