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/518 Thanks! 1 00:03:46,810 --> 00:03:49,119 But as of the end of November, 2 00:03:49,120 --> 00:03:50,649 Rhim had published a blog post saying 3 00:03:50,650 --> 00:03:51,999 that they were expecting to leave. 4 00:03:52,000 --> 00:03:53,759 So we'll see how that how that goes. 5 00:03:56,680 --> 00:03:58,779 Back in East Asia, Korea 6 00:04:00,610 --> 00:04:02,739 has implemented 7 00:04:02,740 --> 00:04:05,379 a thing to protect children called 8 00:04:05,380 --> 00:04:07,569 Smart Sheriff, and 9 00:04:07,570 --> 00:04:08,889 I mean, that's one of the implementations, 10 00:04:08,890 --> 00:04:11,469 but it comes out of a policy to 11 00:04:11,470 --> 00:04:13,119 to add additional filtering. 12 00:04:13,120 --> 00:04:15,189 It's required on sort of mobile devices 13 00:04:15,190 --> 00:04:17,859 need to have this child protection thing. 14 00:04:17,860 --> 00:04:19,958 And there was a report this fall after 15 00:04:19,959 --> 00:04:21,518 the implementation in the spring saying 16 00:04:21,519 --> 00:04:22,779 that it actually opens up a bunch of 17 00:04:22,780 --> 00:04:24,759 security issues and a bunch of 18 00:04:24,760 --> 00:04:26,289 surveillance issues on the devices that 19 00:04:26,290 --> 00:04:27,939 it's on. So there's there's sort of an 20 00:04:27,940 --> 00:04:30,579 ongoing conversation about how to do this 21 00:04:30,580 --> 00:04:33,099 and what the right level of safety 22 00:04:33,100 --> 00:04:34,100 versus. 23 00:04:35,570 --> 00:04:36,770 The other type of safety is. 24 00:04:38,870 --> 00:04:41,119 Not to be outdone, North Korea has added 25 00:04:41,120 --> 00:04:42,120 a squid proxy. 26 00:04:47,360 --> 00:04:50,239 So from the European 27 00:04:50,240 --> 00:04:52,789 segment size, I would like to 28 00:04:52,790 --> 00:04:55,789 mention a new so-called, 29 00:04:55,790 --> 00:04:58,309 as I used to say it senses it and 30 00:04:58,310 --> 00:05:00,709 also known as gambling 31 00:05:00,710 --> 00:05:01,710 censorship. 32 00:05:02,630 --> 00:05:04,849 Apparently the governments have 33 00:05:04,850 --> 00:05:07,189 decided that they need to 34 00:05:07,190 --> 00:05:09,349 better regulate the gambling and betting 35 00:05:09,350 --> 00:05:11,209 resorts on the Internet. 36 00:05:11,210 --> 00:05:13,639 And they have initiated 37 00:05:13,640 --> 00:05:15,979 a sort of putting more 38 00:05:15,980 --> 00:05:18,949 and more regulation and policies to 39 00:05:18,950 --> 00:05:20,480 block, filter and 40 00:05:22,100 --> 00:05:24,179 somehow reduce access to 41 00:05:24,180 --> 00:05:25,180 these resources. 42 00:05:26,300 --> 00:05:29,239 Unfortunately, there are many mistaken 43 00:05:29,240 --> 00:05:30,240 blocking 44 00:05:31,610 --> 00:05:33,439 virus cases of forward blocking 45 00:05:35,000 --> 00:05:37,279 as far as collateral 46 00:05:37,280 --> 00:05:40,489 damage cases like Ecosystem's 47 00:05:40,490 --> 00:05:43,249 DNS, Mellick Saints record 48 00:05:43,250 --> 00:05:45,379 or actually less 49 00:05:45,380 --> 00:05:47,779 time out on websites 50 00:05:47,780 --> 00:05:49,279 that goes on the same IP. 51 00:05:49,280 --> 00:05:51,889 It's quite common in the webspace 52 00:05:51,890 --> 00:05:53,719 environment where actually we said 53 00:05:53,720 --> 00:05:55,879 hosting. You said the same IP 54 00:05:55,880 --> 00:05:57,869 with the rest of the other side. 55 00:05:59,750 --> 00:06:02,149 What I would like to mention 56 00:06:02,150 --> 00:06:04,789 here, it's a report 57 00:06:04,790 --> 00:06:07,249 from an organization 58 00:06:07,250 --> 00:06:08,869 in Switzerland for the European 59 00:06:08,870 --> 00:06:10,189 Commission. 60 00:06:10,190 --> 00:06:12,679 It's like more 61 00:06:12,680 --> 00:06:15,499 than 1500 pages 62 00:06:15,500 --> 00:06:17,329 and it's from 2006. 63 00:06:19,100 --> 00:06:20,689 What is happening actually here? 64 00:06:20,690 --> 00:06:22,279 It's like a table 65 00:06:23,330 --> 00:06:25,819 that mentions the 66 00:06:25,820 --> 00:06:28,069 the counties from left to right. 67 00:06:28,070 --> 00:06:30,349 Like from left are the stricter controls 68 00:06:30,350 --> 00:06:33,139 with policies and gambling 69 00:06:33,140 --> 00:06:35,019 and betting regulation to the right, 70 00:06:35,020 --> 00:06:37,129 which they have less and less 71 00:06:38,390 --> 00:06:40,589 regulation. This has changed already. 72 00:06:40,590 --> 00:06:42,709 So, for instance, UK, on the 73 00:06:42,710 --> 00:06:45,439 far right side, it 74 00:06:45,440 --> 00:06:47,569 actually does some 75 00:06:47,570 --> 00:06:49,669 extreme blocking, which I'm going to be 76 00:06:49,670 --> 00:06:52,190 mentioned later in my slides. 77 00:06:53,930 --> 00:06:56,719 What I would like actually to give 78 00:06:56,720 --> 00:06:58,849 to point out, it's a case 79 00:06:58,850 --> 00:07:00,949 that we did some work 80 00:07:00,950 --> 00:07:01,519 later. 81 00:07:01,520 --> 00:07:05,059 Uh, before this summer 82 00:07:05,060 --> 00:07:06,679 was the case of Greece. 83 00:07:06,680 --> 00:07:08,899 We did some extensive study 84 00:07:08,900 --> 00:07:11,299 on quite many ISPs. 85 00:07:11,300 --> 00:07:14,089 Actually, we cover most of ISPs 86 00:07:14,090 --> 00:07:16,279 and we. 87 00:07:16,280 --> 00:07:18,059 Well, sorry for that. 88 00:07:19,700 --> 00:07:21,889 So what's actually 89 00:07:21,890 --> 00:07:24,169 happening in Greece is that 90 00:07:24,170 --> 00:07:26,269 the ISP, somehow they 91 00:07:26,270 --> 00:07:27,270 have started 92 00:07:29,630 --> 00:07:30,919 blocking at will. 93 00:07:30,920 --> 00:07:33,229 So they just decided 94 00:07:33,230 --> 00:07:35,689 that they need to do this regulation. 95 00:07:35,690 --> 00:07:37,579 Okay, so they were they were forced from 96 00:07:37,580 --> 00:07:39,589 the Gaming Commission to do this. 97 00:07:39,590 --> 00:07:42,079 But as you can see on this table, 98 00:07:42,080 --> 00:07:44,509 actually in this field, you are covering, 99 00:07:44,510 --> 00:07:46,999 like most Greek ISPs, 100 00:07:47,000 --> 00:07:49,129 they have have inconsistency in 101 00:07:49,130 --> 00:07:50,029 blocking. 102 00:07:50,030 --> 00:07:52,819 Uh, so they're 103 00:07:52,820 --> 00:07:55,069 there to represent the site that has 104 00:07:55,070 --> 00:07:57,889 been blocked, actually, that far right 105 00:07:57,890 --> 00:08:00,019 column. It's like the total side on this 106 00:08:00,020 --> 00:08:02,419 snapshot of the blacklist that should 107 00:08:02,420 --> 00:08:04,939 have been blocked and 108 00:08:04,940 --> 00:08:06,349 the green was this. 109 00:08:06,350 --> 00:08:08,719 The result is the end result 110 00:08:08,720 --> 00:08:10,729 hasn't been blocked yet. 111 00:08:10,730 --> 00:08:13,399 Uh, I can't cover 112 00:08:13,400 --> 00:08:15,709 the whole research about 113 00:08:15,710 --> 00:08:17,269 Greece, but I'm going to give some 114 00:08:17,270 --> 00:08:19,459 examples of what 115 00:08:19,460 --> 00:08:21,319 have happened after this and what was 116 00:08:21,320 --> 00:08:22,499 happening during the. 117 00:08:22,500 --> 00:08:24,679 So mostly they had done like DNS 118 00:08:24,680 --> 00:08:25,680 blocking. 119 00:08:27,320 --> 00:08:29,779 Some companies decided 120 00:08:29,780 --> 00:08:32,449 to just don't, uh, 121 00:08:32,450 --> 00:08:33,829 administer this well. 122 00:08:33,830 --> 00:08:36,229 And they have issued like eighty 123 00:08:36,230 --> 00:08:37,729 four hundred three errors. 124 00:08:37,730 --> 00:08:40,099 And in some cases like 125 00:08:40,100 --> 00:08:41,570 Vodafone, some of them, 126 00:08:42,770 --> 00:08:45,469 they have done some 127 00:08:45,470 --> 00:08:46,470 VPI. 128 00:08:47,610 --> 00:08:49,849 Uh, so what's uh what 129 00:08:49,850 --> 00:08:50,850 the uses were 130 00:08:52,400 --> 00:08:53,640 used to uh. 131 00:08:55,540 --> 00:08:57,889 Experience actually was that this 132 00:08:57,890 --> 00:09:00,009 blocking page, this is this is 133 00:09:00,010 --> 00:09:01,569 quite important if you consider that 134 00:09:01,570 --> 00:09:03,519 Greece was considered to be too hard, 135 00:09:03,520 --> 00:09:04,629 like quite liberal 136 00:09:05,800 --> 00:09:07,989 views on the Internet, you know, like you 137 00:09:07,990 --> 00:09:10,119 still have a 138 00:09:10,120 --> 00:09:12,369 can download some 139 00:09:12,370 --> 00:09:14,829 torrents. And you can I mean, 140 00:09:14,830 --> 00:09:16,779 there is there was some IP blocking or 141 00:09:16,780 --> 00:09:18,699 there were some defamations. 142 00:09:18,700 --> 00:09:20,709 But in other words, there was not such 143 00:09:20,710 --> 00:09:22,209 extensive filtering. 144 00:09:22,210 --> 00:09:24,339 So most of the of the 145 00:09:24,340 --> 00:09:25,960 ISPs were making like. 146 00:09:27,080 --> 00:09:28,940 It's like this one that you see right now 147 00:09:30,140 --> 00:09:32,389 where the other users were being 148 00:09:32,390 --> 00:09:34,459 landed. What is actually 149 00:09:34,460 --> 00:09:36,919 happening now is 150 00:09:36,920 --> 00:09:39,439 that the Espy's are 151 00:09:39,440 --> 00:09:41,989 redirecting them to 152 00:09:41,990 --> 00:09:44,659 the Gaming Commission website. 153 00:09:44,660 --> 00:09:46,909 A reminder to the Gaming Commission 154 00:09:46,910 --> 00:09:49,369 is actually the commission that is 155 00:09:49,370 --> 00:09:51,469 the blacklist, forcing the blacklist 156 00:09:51,470 --> 00:09:53,299 and making laws. 157 00:09:53,300 --> 00:09:55,429 One of them is to punish or 158 00:09:55,430 --> 00:09:58,909 like it's considered 159 00:09:58,910 --> 00:10:01,159 a crime to even access 160 00:10:01,160 --> 00:10:02,570 these forbidden 161 00:10:03,710 --> 00:10:05,149 web resources. 162 00:10:05,150 --> 00:10:07,999 So the users are actually experience 163 00:10:08,000 --> 00:10:10,489 such a block right now with these 164 00:10:10,490 --> 00:10:12,979 these generic emad's 165 00:10:12,980 --> 00:10:15,919 from custom new 166 00:10:15,920 --> 00:10:18,889 data center binding that everything's 167 00:10:18,890 --> 00:10:19,890 fine. 168 00:10:21,680 --> 00:10:24,079 What is quite interesting 169 00:10:24,080 --> 00:10:26,269 to mention and hasn't been mentioned yet, 170 00:10:28,430 --> 00:10:30,679 and I think it makes sense to 171 00:10:30,680 --> 00:10:32,360 something at some point and even. 172 00:10:33,480 --> 00:10:35,729 So this right now is 173 00:10:35,730 --> 00:10:38,069 that since the Greek 174 00:10:38,070 --> 00:10:40,139 escapees implemented 175 00:10:40,140 --> 00:10:41,140 these. 176 00:10:42,200 --> 00:10:43,750 Filtering infrastructure, 177 00:10:44,780 --> 00:10:47,059 they have started actually using the same 178 00:10:47,060 --> 00:10:49,249 infrastructure to block different 179 00:10:49,250 --> 00:10:51,769 resources, lots of so-called 180 00:10:51,770 --> 00:10:53,989 illegal content, as you can see, 181 00:10:53,990 --> 00:10:56,269 that is encoded domain 182 00:10:56,270 --> 00:10:58,069 name, it's in Greek. 183 00:10:58,070 --> 00:11:00,259 So for those of the other 184 00:11:00,260 --> 00:11:02,329 candidates TV, it's like it's quite 185 00:11:02,330 --> 00:11:03,319 simple. 186 00:11:03,320 --> 00:11:05,629 The body, it's it's not 187 00:11:05,630 --> 00:11:06,769 it's there's nothing there. 188 00:11:06,770 --> 00:11:09,349 And it's returning there to be status 189 00:11:09,350 --> 00:11:11,539 for zero three, no Knoblock 190 00:11:11,540 --> 00:11:12,469 page, no nothing. 191 00:11:12,470 --> 00:11:14,899 So and this 192 00:11:14,900 --> 00:11:17,509 has been done by a vote of this 193 00:11:17,510 --> 00:11:18,809 specific segment. 194 00:11:18,810 --> 00:11:20,240 Has it done by Vodafone 195 00:11:21,590 --> 00:11:23,759 in June 2015 196 00:11:23,760 --> 00:11:26,399 that they have used the local 197 00:11:26,400 --> 00:11:28,879 proxy for 198 00:11:28,880 --> 00:11:30,349 utilizing DPI 199 00:11:31,430 --> 00:11:32,430 of. 200 00:11:33,770 --> 00:11:36,079 Moving on, another interesting 201 00:11:36,080 --> 00:11:38,389 case, United Kingdom, 202 00:11:39,950 --> 00:11:42,199 we actually have two presentations for 203 00:11:42,200 --> 00:11:45,049 United Kingdom, but yeah, that 204 00:11:45,050 --> 00:11:47,090 have only one and we could 205 00:11:48,170 --> 00:11:50,209 not mention so many things that happened 206 00:11:50,210 --> 00:11:53,119 the last years, but 207 00:11:53,120 --> 00:11:54,889 something quite important happened in 208 00:11:54,890 --> 00:11:57,169 2005 or and was 209 00:11:57,170 --> 00:11:59,389 increased in intensity 210 00:11:59,390 --> 00:12:01,219 in 2013, that the government 211 00:12:02,240 --> 00:12:04,459 has started pressuring 212 00:12:04,460 --> 00:12:06,889 a lot for the of promoting 213 00:12:06,890 --> 00:12:08,899 filters to Espy's. 214 00:12:08,900 --> 00:12:12,019 But in the difference with 215 00:12:12,020 --> 00:12:14,119 UK that they didn't they 216 00:12:14,120 --> 00:12:15,949 don't really have a gaming commission or 217 00:12:15,950 --> 00:12:18,109 they don't have like a commission that 218 00:12:18,110 --> 00:12:20,059 makes what should be blocked and, you 219 00:12:20,060 --> 00:12:22,009 know, makes a blacklist and whatever. 220 00:12:22,010 --> 00:12:24,289 So it's ISP decide on what 221 00:12:24,290 --> 00:12:26,479 is right that needs 222 00:12:26,480 --> 00:12:28,819 to be blocked, filter and 223 00:12:28,820 --> 00:12:31,009 censor so you can actually find out 224 00:12:31,010 --> 00:12:33,199 what happened to like, you know, like 225 00:12:33,200 --> 00:12:35,179 dentist websites have been blocked, some 226 00:12:35,180 --> 00:12:38,029 bars have been blocked or some 227 00:12:38,030 --> 00:12:40,109 websites that used to do like to do 228 00:12:40,110 --> 00:12:42,499 a forum like the the keyword 229 00:12:42,500 --> 00:12:43,879 six have been blocked. 230 00:12:43,880 --> 00:12:45,710 So a lot of false positives 231 00:12:47,120 --> 00:12:49,399 are what is 232 00:12:50,930 --> 00:12:53,089 quite important. Medicine is 233 00:12:53,090 --> 00:12:55,249 Blocked Project from Open Rights 234 00:12:55,250 --> 00:12:56,250 Group. 235 00:12:57,110 --> 00:12:59,929 They have done some impressive work 236 00:12:59,930 --> 00:13:01,249 monitoring all this. 237 00:13:01,250 --> 00:13:03,409 Uh, like all I, Espy's or 238 00:13:03,410 --> 00:13:05,659 most ISPs in, uh, 239 00:13:05,660 --> 00:13:07,729 U.K., uh, they 240 00:13:07,730 --> 00:13:09,859 have used their legs up to 241 00:13:09,860 --> 00:13:12,539 K and these are 242 00:13:12,540 --> 00:13:14,689 results like are from there 243 00:13:14,690 --> 00:13:16,579 yesterday snapshot. 244 00:13:16,580 --> 00:13:18,919 So they have actually, uh, find 245 00:13:18,920 --> 00:13:20,869 out that they're are around 20 from these 246 00:13:20,870 --> 00:13:23,599 hundred gather around twenty 247 00:13:23,600 --> 00:13:25,440 twenty one thousand, uh, 248 00:13:26,450 --> 00:13:28,609 websites. Being a resource is being 249 00:13:28,610 --> 00:13:30,889 blocked by strict filters and 250 00:13:30,890 --> 00:13:33,349 around 11000 websites 251 00:13:33,350 --> 00:13:35,059 block by default filters. 252 00:13:35,060 --> 00:13:37,039 Let me get clear, what is default 253 00:13:37,040 --> 00:13:39,139 filters? So when you buy a new one, 254 00:13:39,140 --> 00:13:41,209 you are becoming a subscriber or you 255 00:13:41,210 --> 00:13:43,889 buy a new cellular number in U.K., 256 00:13:43,890 --> 00:13:45,619 you're coming with this default filter. 257 00:13:45,620 --> 00:13:47,329 So you actually need to opt out. 258 00:13:47,330 --> 00:13:49,039 You need to say, hey, you know, I want to 259 00:13:49,040 --> 00:13:51,229 watch porn. I want to have like 260 00:13:51,230 --> 00:13:53,479 an uncensored Internet 261 00:13:53,480 --> 00:13:54,859 connection. Uh. 262 00:13:57,010 --> 00:13:59,379 And this is like the latest 263 00:13:59,380 --> 00:14:01,779 stop of the yesterday snapshots 264 00:14:01,780 --> 00:14:02,739 on their page. 265 00:14:02,740 --> 00:14:04,569 You can actually click on the slides, 266 00:14:04,570 --> 00:14:06,779 which we're going to publish online. 267 00:14:08,660 --> 00:14:10,839 They're actually having their 268 00:14:10,840 --> 00:14:13,239 provider, the presenters 269 00:14:13,240 --> 00:14:16,179 of sites that have me blog and 270 00:14:16,180 --> 00:14:17,289 some discrepencies. 271 00:14:21,160 --> 00:14:23,229 Hmm, so I 272 00:14:23,230 --> 00:14:26,079 would like to move to a different 273 00:14:26,080 --> 00:14:27,639 segment right now. 274 00:14:27,640 --> 00:14:29,890 This is the Middle East 275 00:14:32,260 --> 00:14:34,659 and continue with Saudi 276 00:14:34,660 --> 00:14:35,660 Arabia. 277 00:14:36,700 --> 00:14:39,669 This is an exit forum and 278 00:14:39,670 --> 00:14:41,799 only a report from 279 00:14:41,800 --> 00:14:42,879 an active request. 280 00:14:42,880 --> 00:14:45,100 This I will explain more later. 281 00:14:46,270 --> 00:14:48,219 What is happening actually here in the 282 00:14:48,220 --> 00:14:50,079 case of Saudi Arabia is that they have 283 00:14:50,080 --> 00:14:51,759 actually blocked the Arab times. 284 00:14:51,760 --> 00:14:55,179 This happened like this is 2015. 285 00:14:55,180 --> 00:14:57,729 And they have just, uh, 286 00:14:57,730 --> 00:15:00,489 brought like a random I just, 287 00:15:00,490 --> 00:15:03,499 uh, saw this page blocked 288 00:15:03,500 --> 00:15:06,579 with an indication of this, uh, 289 00:15:06,580 --> 00:15:09,189 uh, resource having been blocked by 290 00:15:09,190 --> 00:15:10,489 Widefield. 291 00:15:10,490 --> 00:15:13,119 Uh, I did some 292 00:15:13,120 --> 00:15:15,309 search and I found out that that's 293 00:15:15,310 --> 00:15:16,719 the wire filter. 294 00:15:16,720 --> 00:15:19,779 It seems like a Web proxy, GPI, whatever, 295 00:15:19,780 --> 00:15:23,289 with, you know, uh, supposedly 296 00:15:23,290 --> 00:15:26,379 on their website, uh, you should be used 297 00:15:26,380 --> 00:15:27,639 for good ways. 298 00:15:27,640 --> 00:15:29,829 And so, you know, for, as 299 00:15:29,830 --> 00:15:32,079 they call it, for a worry free 300 00:15:32,080 --> 00:15:33,840 Internet experiences. 301 00:15:38,410 --> 00:15:39,700 Continuing with Iraq. 302 00:15:41,020 --> 00:15:43,239 Does anyone know what happened on to on 303 00:15:43,240 --> 00:15:46,209 the normally happens end of July 304 00:15:46,210 --> 00:15:48,549 or specifically on twenty seventh 305 00:15:48,550 --> 00:15:50,290 of July in 2015? 306 00:15:54,080 --> 00:15:56,329 OK, so 307 00:15:56,330 --> 00:15:58,819 on this, during this time, 308 00:15:58,820 --> 00:16:01,309 normally there, the 309 00:16:01,310 --> 00:16:03,859 school, the student class exams 310 00:16:03,860 --> 00:16:05,359 in Iraq 311 00:16:06,620 --> 00:16:08,779 and what would 312 00:16:08,780 --> 00:16:10,969 happen actually between 313 00:16:10,970 --> 00:16:13,429 two to five, like seven 314 00:16:13,430 --> 00:16:15,679 to 10 local time is that, 315 00:16:15,680 --> 00:16:19,279 as you can see from this BTP dataflow, 316 00:16:19,280 --> 00:16:22,789 is that Iraq decided to actually 317 00:16:22,790 --> 00:16:24,439 stop a bit saying they're not, you know, 318 00:16:24,440 --> 00:16:26,569 for the students to have to 319 00:16:26,570 --> 00:16:28,849 worry free exams without teaching. 320 00:16:28,850 --> 00:16:29,850 And then we're. 321 00:16:32,260 --> 00:16:34,059 We moved on with our exam's. 322 00:16:42,780 --> 00:16:45,869 Another interesting report that we found 323 00:16:45,870 --> 00:16:48,119 by scaping their only 324 00:16:48,120 --> 00:16:50,189 report was the 325 00:16:50,190 --> 00:16:52,709 HTP involved request Linowitz, 326 00:16:52,710 --> 00:16:55,829 so that there were some instances of 327 00:16:55,830 --> 00:16:57,929 squit proxy and 328 00:16:57,930 --> 00:17:00,210 proxy that has been used from 329 00:17:01,620 --> 00:17:03,839 the Mensing here, from this Firebaugh, 330 00:17:03,840 --> 00:17:05,280 from Vod's lying. 331 00:17:06,650 --> 00:17:09,279 I don't know what they do, uh, 332 00:17:10,369 --> 00:17:13,249 moving on with Iran and 333 00:17:13,250 --> 00:17:15,409 specifically a very 334 00:17:15,410 --> 00:17:17,719 small segment of what happened 335 00:17:17,720 --> 00:17:19,789 in Iran, as I mentioned, that 336 00:17:19,790 --> 00:17:21,348 it's a very short presentation. 337 00:17:21,349 --> 00:17:23,539 We need to cover a lot of 338 00:17:23,540 --> 00:17:24,540 things. 339 00:17:25,280 --> 00:17:27,949 So Iran decided to actually 340 00:17:27,950 --> 00:17:30,409 they have been blocking telegram 341 00:17:30,410 --> 00:17:32,659 and so many other things about. 342 00:17:32,660 --> 00:17:34,919 What's interesting in this case is that 343 00:17:34,920 --> 00:17:36,159 on No. 344 00:17:36,160 --> 00:17:37,640 20 twenties, 345 00:17:39,230 --> 00:17:42,020 they were their users were starting 346 00:17:43,040 --> 00:17:44,359 wondering social 347 00:17:45,590 --> 00:17:47,809 media sites like, hey, what's happening? 348 00:17:47,810 --> 00:17:48,709 Is it blocked or not? 349 00:17:48,710 --> 00:17:50,779 And then said, no, actually, it wasn't 350 00:17:50,780 --> 00:17:51,859 blocked. There were some 351 00:17:53,090 --> 00:17:55,630 disturbances that are 352 00:17:57,350 --> 00:17:59,839 technical issues that help us, too. 353 00:17:59,840 --> 00:18:01,429 That didn't help. 354 00:18:01,430 --> 00:18:03,949 And we had somehow 355 00:18:03,950 --> 00:18:05,369 this. 356 00:18:05,370 --> 00:18:07,649 Brought this side of line, 357 00:18:07,650 --> 00:18:09,540 this telegram offline, 358 00:18:13,230 --> 00:18:15,239 and now I'm giving the waterwheel. 359 00:18:16,750 --> 00:18:18,899 Cool, so, 360 00:18:18,900 --> 00:18:20,550 yeah, so in the. 361 00:18:22,200 --> 00:18:23,759 West, one of the main things that's 362 00:18:23,760 --> 00:18:25,229 happened is trying to understand how to 363 00:18:25,230 --> 00:18:28,139 deal with this rise of 364 00:18:28,140 --> 00:18:30,029 Daesh or ISIL or whatever we're going to 365 00:18:30,030 --> 00:18:31,030 call it at the moment. 366 00:18:32,460 --> 00:18:35,549 And there's sort of this recognition 367 00:18:35,550 --> 00:18:37,949 in politics that, you know, 368 00:18:37,950 --> 00:18:39,209 suddenly there's bad stuff on the 369 00:18:39,210 --> 00:18:41,879 Internet again and we need to, you know, 370 00:18:41,880 --> 00:18:43,119 do do something about it. 371 00:18:43,120 --> 00:18:44,489 And lest we think it's, you know, from 372 00:18:44,490 --> 00:18:46,619 one side, this is sort of a thing that's 373 00:18:46,620 --> 00:18:47,620 sort of generally 374 00:18:48,900 --> 00:18:51,479 in the political mindset right now of, 375 00:18:51,480 --> 00:18:53,669 well, you know, it's not censorship if if 376 00:18:53,670 --> 00:18:55,439 Twitter, you know, solves this problem 377 00:18:55,440 --> 00:18:56,909 and maybe we should talk to Bill Gates 378 00:18:56,910 --> 00:18:58,320 and he'll close the Internet for us. 379 00:19:00,690 --> 00:19:02,759 And there sort of seems to actually 380 00:19:02,760 --> 00:19:04,889 be this response from 381 00:19:04,890 --> 00:19:07,109 Silicon Valley and these big companies 382 00:19:07,110 --> 00:19:09,389 to try and police their content more. 383 00:19:09,390 --> 00:19:11,459 Right. And we've seen also that as 384 00:19:11,460 --> 00:19:13,979 the safe harbor agreement with the EU 385 00:19:13,980 --> 00:19:16,199 sort of fell apart over the summer, that, 386 00:19:16,200 --> 00:19:17,909 again, all of these content providers are 387 00:19:17,910 --> 00:19:20,009 taking a closer look at management 388 00:19:20,010 --> 00:19:21,779 of their their networks and their user 389 00:19:21,780 --> 00:19:23,069 generated content. 390 00:19:23,070 --> 00:19:25,229 And so you're seeing things like Twitter 391 00:19:25,230 --> 00:19:27,179 is suspending the ISIS affiliated 392 00:19:27,180 --> 00:19:29,579 accounts and similar things 393 00:19:29,580 --> 00:19:31,509 on like Facebook or whatever. 394 00:19:31,510 --> 00:19:32,839 Right. 395 00:19:32,840 --> 00:19:34,519 And so this is actually showing this big 396 00:19:34,520 --> 00:19:36,709 hole that we've got, which is many 397 00:19:36,710 --> 00:19:38,269 of these companies have transparency 398 00:19:38,270 --> 00:19:39,809 reports. 399 00:19:39,810 --> 00:19:41,369 We can see when the government is asking 400 00:19:41,370 --> 00:19:43,349 for subscriber information and we can see 401 00:19:43,350 --> 00:19:45,689 when they ask for explicit content 402 00:19:45,690 --> 00:19:47,909 to be removed and we can see what other 403 00:19:47,910 --> 00:19:50,199 companies make copyright notices. 404 00:19:50,200 --> 00:19:51,389 But there's really nothing here about how 405 00:19:51,390 --> 00:19:53,579 much is being taken down by Twitter 406 00:19:53,580 --> 00:19:54,929 itself. 407 00:19:54,930 --> 00:19:56,399 And so we don't have great insight. 408 00:19:56,400 --> 00:19:58,199 And so taking this full circle, we 409 00:19:58,200 --> 00:20:00,569 actually do have these technologies. 410 00:20:00,570 --> 00:20:02,729 We use them for sites like Weibo, 411 00:20:02,730 --> 00:20:05,219 the Chinese equivalent to Twitter, where 412 00:20:05,220 --> 00:20:06,929 this open source project called Freebo 413 00:20:06,930 --> 00:20:09,119 Weibo monitors 414 00:20:09,120 --> 00:20:11,219 the Chinese social network finds popular 415 00:20:11,220 --> 00:20:12,809 content that then is subsequently removed 416 00:20:12,810 --> 00:20:14,190 and reposted for posterity. 417 00:20:15,380 --> 00:20:17,029 And so this is open source, I'd recommend 418 00:20:17,030 --> 00:20:18,919 that someone do this for some of the 419 00:20:18,920 --> 00:20:21,169 Western media as well so that we can see 420 00:20:21,170 --> 00:20:22,170 what's happening. 421 00:20:22,820 --> 00:20:24,979 And I guess it's worth also 422 00:20:24,980 --> 00:20:26,509 pointing out that a lot of this stuff is 423 00:20:26,510 --> 00:20:29,029 happening on totally public websites. 424 00:20:29,030 --> 00:20:30,769 Right. As we get into Facebook and 425 00:20:30,770 --> 00:20:32,659 WhatsApp groups and telegram and all of 426 00:20:32,660 --> 00:20:34,219 these content providers that are having 427 00:20:34,220 --> 00:20:35,839 to wrestle with this, a lot of this 428 00:20:35,840 --> 00:20:36,769 content is not public. 429 00:20:36,770 --> 00:20:38,359 It's within small groups. 430 00:20:38,360 --> 00:20:39,829 And so it's going to be very hard for us 431 00:20:39,830 --> 00:20:41,629 to understand what sort of management 432 00:20:41,630 --> 00:20:43,609 policies are happening on what's going 433 00:20:43,610 --> 00:20:44,610 on. 434 00:20:45,910 --> 00:20:47,979 So that's sort of our whirlwind tour 435 00:20:47,980 --> 00:20:50,079 of sort of things that 436 00:20:50,080 --> 00:20:51,219 we thought were worth mentioning, there's 437 00:20:51,220 --> 00:20:52,599 tons more. 438 00:20:52,600 --> 00:20:54,609 And what we want to talk about in the in 439 00:20:54,610 --> 00:20:57,189 sort of the second half of this is 440 00:20:57,190 --> 00:20:58,479 how do we know that? 441 00:20:58,480 --> 00:21:00,579 What are the projects and tools 442 00:21:00,580 --> 00:21:02,829 that we've got that help us 443 00:21:02,830 --> 00:21:05,079 see when 444 00:21:05,080 --> 00:21:07,179 there are disturbances and 445 00:21:07,180 --> 00:21:08,490 what perspectives we can get? 446 00:21:09,670 --> 00:21:11,020 So we mentioned that the. 447 00:21:12,570 --> 00:21:14,709 Initial map, and 448 00:21:14,710 --> 00:21:16,739 that comes from this group called Freedom 449 00:21:16,740 --> 00:21:18,209 House, and they publish a report each 450 00:21:18,210 --> 00:21:19,679 year called Freedom on the Net. 451 00:21:19,680 --> 00:21:22,249 And what that looks like is they 452 00:21:22,250 --> 00:21:24,329 they write up a a long sort 453 00:21:24,330 --> 00:21:26,489 of narrative for each country and 454 00:21:26,490 --> 00:21:29,519 then they have analysts code 455 00:21:29,520 --> 00:21:31,859 free, somewhat free, not free against 456 00:21:31,860 --> 00:21:32,860 that. 457 00:21:33,340 --> 00:21:34,589 And so they do this. 458 00:21:34,590 --> 00:21:35,939 And there's another group called Open Net 459 00:21:35,940 --> 00:21:37,889 Initiative that also does this. 460 00:21:37,890 --> 00:21:40,469 And so you get maybe a few categories 461 00:21:40,470 --> 00:21:42,869 like how free is it on 462 00:21:42,870 --> 00:21:44,069 the political spectrum? 463 00:21:44,070 --> 00:21:46,589 How free is it on, you know, access 464 00:21:46,590 --> 00:21:47,910 to circumvention tools? 465 00:21:49,850 --> 00:21:51,019 But what you don't get is you don't get 466 00:21:51,020 --> 00:21:52,489 the data backing that you don't get the 467 00:21:52,490 --> 00:21:54,469 list of specific sites you don't get when 468 00:21:54,470 --> 00:21:56,479 things were blocked because all of that 469 00:21:56,480 --> 00:21:57,479 is potentially harmful. 470 00:21:57,480 --> 00:21:59,599 Right. If I if I publish all of the 471 00:21:59,600 --> 00:22:01,819 data that I'm using to make my 472 00:22:01,820 --> 00:22:04,249 analysis, whoever is managing 473 00:22:04,250 --> 00:22:05,779 that network can potentially go back and 474 00:22:05,780 --> 00:22:07,159 figure out who it was, who ran those 475 00:22:07,160 --> 00:22:09,139 measurements and potentially say, hey, 476 00:22:09,140 --> 00:22:10,140 knock it off. 477 00:22:11,860 --> 00:22:12,860 So. 478 00:22:14,110 --> 00:22:15,369 This is this is something that's been 479 00:22:15,370 --> 00:22:17,199 sort of struggled with for a while now by 480 00:22:17,200 --> 00:22:18,159 the measurement community, and we're 481 00:22:18,160 --> 00:22:19,160 making progress. 482 00:22:19,960 --> 00:22:21,609 And so we're going to start with sort of 483 00:22:21,610 --> 00:22:24,009 the the big player in the 484 00:22:24,010 --> 00:22:25,010 in the room. 485 00:22:25,840 --> 00:22:27,859 So, yeah. How we measure interference 486 00:22:27,860 --> 00:22:28,860 then. 487 00:22:32,080 --> 00:22:34,179 Only disclosure, 488 00:22:34,180 --> 00:22:35,440 I'm an only developer. 489 00:22:37,300 --> 00:22:39,579 Also known as open observatory 490 00:22:39,580 --> 00:22:40,749 network interference. 491 00:22:42,460 --> 00:22:44,959 It's been around for some years. 492 00:22:44,960 --> 00:22:46,869 It has some peer reviewed 493 00:22:47,980 --> 00:22:48,989 proof of concept. 494 00:22:50,020 --> 00:22:52,329 It has been used to 495 00:22:52,330 --> 00:22:54,699 analyze in general, 496 00:22:54,700 --> 00:22:56,799 like most of the things that I 497 00:22:56,800 --> 00:22:59,289 saw during my presentation. 498 00:22:59,290 --> 00:23:00,290 Uh. 499 00:23:01,530 --> 00:23:03,750 It's free software, 500 00:23:05,160 --> 00:23:07,589 you could tell, 501 00:23:07,590 --> 00:23:09,989 by adding a specific test models. 502 00:23:12,120 --> 00:23:14,309 Uh, long story short, 503 00:23:14,310 --> 00:23:16,889 I can point you to somehow 504 00:23:16,890 --> 00:23:19,399 how this looks like, uh, 505 00:23:19,400 --> 00:23:20,309 weight. 506 00:23:20,310 --> 00:23:22,619 It's a measurement 507 00:23:22,620 --> 00:23:24,899 platform that has a back 508 00:23:24,900 --> 00:23:27,179 end, has parts of 509 00:23:27,180 --> 00:23:29,369 clients that are submitting their 510 00:23:29,370 --> 00:23:30,399 reports. 511 00:23:30,400 --> 00:23:32,979 Uh, Brian, 512 00:23:32,980 --> 00:23:35,879 a new signing pipeline is 513 00:23:35,880 --> 00:23:37,439 being developed. 514 00:23:37,440 --> 00:23:39,539 There is an API that is going 515 00:23:39,540 --> 00:23:41,669 to be online 516 00:23:41,670 --> 00:23:44,759 soon and give users 517 00:23:44,760 --> 00:23:47,159 the ability to search in a visual 518 00:23:47,160 --> 00:23:49,949 way and, uh, 519 00:23:49,950 --> 00:23:52,109 sysadmin part that is going to 520 00:23:52,110 --> 00:23:54,599 Meynard and do all this stuff 521 00:23:54,600 --> 00:23:56,489 like storing the reports. 522 00:23:56,490 --> 00:23:58,499 Because actually what is happening with 523 00:23:58,500 --> 00:24:01,379 is that it stores the reports publicly, 524 00:24:01,380 --> 00:24:03,929 so allows all people, 525 00:24:03,930 --> 00:24:06,479 researchers or 526 00:24:06,480 --> 00:24:09,179 interesting users to search 527 00:24:09,180 --> 00:24:12,119 and somehow 528 00:24:12,120 --> 00:24:13,650 start exploring 529 00:24:15,030 --> 00:24:16,679 through the reports, through the only 530 00:24:16,680 --> 00:24:17,680 report. 531 00:24:19,350 --> 00:24:21,869 A much simpler approach of 532 00:24:21,870 --> 00:24:23,939 a test of a test mission 533 00:24:23,940 --> 00:24:25,619 in the Saudi Arabia case. 534 00:24:25,620 --> 00:24:27,890 It is a HTP request. 535 00:24:29,190 --> 00:24:31,289 I can't I don't have the time to go 536 00:24:31,290 --> 00:24:32,669 into detail onto that. 537 00:24:32,670 --> 00:24:34,919 But I'll give you only a sort of idea 538 00:24:34,920 --> 00:24:36,089 of how this works. 539 00:24:36,090 --> 00:24:38,639 Actually, it's just sent to HTP 540 00:24:38,640 --> 00:24:40,739 get request to the resource that 541 00:24:40,740 --> 00:24:43,079 we like to probe. 542 00:24:43,080 --> 00:24:45,299 One comes through our connection, 543 00:24:45,300 --> 00:24:47,399 which is potentially tampered 544 00:24:47,400 --> 00:24:49,499 or filters or whatever, and 545 00:24:49,500 --> 00:24:51,989 the other one comes from another. 546 00:24:55,620 --> 00:24:57,869 Connection like from door, 547 00:24:57,870 --> 00:25:00,119 and then there is some kind of 548 00:25:00,120 --> 00:25:02,599 deep happening and, uh, 549 00:25:02,600 --> 00:25:04,679 uh, some other 550 00:25:04,680 --> 00:25:07,019 types of, uh, 551 00:25:07,020 --> 00:25:08,999 customizations that are being checked in 552 00:25:09,000 --> 00:25:11,309 order to see if our this resource 553 00:25:11,310 --> 00:25:13,119 has been blocked or not. 554 00:25:13,120 --> 00:25:14,120 Um. 555 00:25:15,810 --> 00:25:17,669 There are many, many more things to say, 556 00:25:17,670 --> 00:25:19,859 but, you know, please come and find 557 00:25:19,860 --> 00:25:22,409 me later if you need more or read 558 00:25:22,410 --> 00:25:24,479 on the website and help if you 559 00:25:24,480 --> 00:25:25,379 want. 560 00:25:25,380 --> 00:25:28,259 Another thing that happened, uh, 561 00:25:28,260 --> 00:25:30,359 and that has been working the last year 562 00:25:30,360 --> 00:25:32,939 was, um, a Raspberry 563 00:25:32,940 --> 00:25:35,399 Pi distribution to increase 564 00:25:35,400 --> 00:25:37,919 the Vinton's points of only and 565 00:25:37,920 --> 00:25:40,529 provides more, uh, reports 566 00:25:40,530 --> 00:25:42,719 from different segments 567 00:25:42,720 --> 00:25:43,809 of the world. 568 00:25:43,810 --> 00:25:45,869 Uh, and in an easy 569 00:25:45,870 --> 00:25:48,179 way. This is Lepidoptera. 570 00:25:48,180 --> 00:25:50,729 Uh, and, uh. 571 00:25:52,490 --> 00:25:54,589 Moving on with I love, it's 572 00:25:56,030 --> 00:25:58,339 a project that has been developed 573 00:25:58,340 --> 00:26:01,099 from the University 574 00:26:01,100 --> 00:26:02,689 of Stony Brook. 575 00:26:02,690 --> 00:26:03,379 Stony Brook. 576 00:26:03,380 --> 00:26:05,340 Yeah, I can't even pronounce the name. 577 00:26:06,570 --> 00:26:08,929 So, yeah, they have 578 00:26:08,930 --> 00:26:11,149 done sort of a closed system of 579 00:26:11,150 --> 00:26:13,459 containing making 580 00:26:13,460 --> 00:26:14,460 measurements. 581 00:26:17,130 --> 00:26:18,890 Uh, another 582 00:26:19,940 --> 00:26:22,069 nice example of collecting 583 00:26:22,070 --> 00:26:25,549 that measurements, but unfortunately 584 00:26:25,550 --> 00:26:28,279 not sensitive 585 00:26:28,280 --> 00:26:30,409 or not 586 00:26:30,410 --> 00:26:32,089 censorship oriented 587 00:26:33,320 --> 00:26:35,599 network measurements collection is that 588 00:26:35,600 --> 00:26:36,600 Atlas 589 00:26:37,850 --> 00:26:39,979 Battler's has a bunch of notes out 590 00:26:39,980 --> 00:26:42,089 there but is being fanned 591 00:26:42,090 --> 00:26:43,609 by the IPCC. 592 00:26:43,610 --> 00:26:45,739 And as you know, or IPCC 593 00:26:45,740 --> 00:26:47,839 has some members like Turkey 594 00:26:47,840 --> 00:26:49,969 and some other organizations that they 595 00:26:49,970 --> 00:26:52,189 would not like to be 596 00:26:52,190 --> 00:26:54,769 associated or to 597 00:26:54,770 --> 00:26:57,170 have any reports that are 598 00:26:58,220 --> 00:26:59,900 negative to their sense. 599 00:27:02,150 --> 00:27:05,209 So, yeah, so in addition to these sort of 600 00:27:05,210 --> 00:27:06,589 distributed platforms. 601 00:27:06,590 --> 00:27:08,659 Right. And so this is a few examples of 602 00:27:08,660 --> 00:27:10,849 places where you've got probes in 603 00:27:10,850 --> 00:27:12,169 networks. 604 00:27:12,170 --> 00:27:14,419 We can learn a lot of stuff by 605 00:27:14,420 --> 00:27:16,069 looking at specific domains. 606 00:27:17,180 --> 00:27:19,699 So there's platforms like Tor 607 00:27:19,700 --> 00:27:21,109 that will provide data. 608 00:27:21,110 --> 00:27:23,179 So on metrics, Dottore, I can look at how 609 00:27:23,180 --> 00:27:25,309 many users are connecting from 610 00:27:25,310 --> 00:27:27,439 Bangladesh and I can verify 611 00:27:27,440 --> 00:27:29,479 that report that we showed earlier that 612 00:27:29,480 --> 00:27:31,579 there was something weird going on from 613 00:27:31,580 --> 00:27:34,009 November to December because 614 00:27:34,010 --> 00:27:35,629 while Facebook was blocked, we see a 615 00:27:35,630 --> 00:27:37,579 corresponding spike in the number of Tor 616 00:27:37,580 --> 00:27:38,749 users. 617 00:27:38,750 --> 00:27:40,909 Right. So they are, you know, providing 618 00:27:40,910 --> 00:27:42,679 this data. And likewise, there's a few 619 00:27:42,680 --> 00:27:44,839 other services like Google 620 00:27:44,840 --> 00:27:46,489 shows how much connections it's getting 621 00:27:46,490 --> 00:27:47,359 from different countries. 622 00:27:47,360 --> 00:27:49,649 And we can say they actually label 623 00:27:49,650 --> 00:27:51,229 themselves and say we believe that there 624 00:27:51,230 --> 00:27:53,479 was a disturbance in these places 625 00:27:53,480 --> 00:27:54,889 at these times. 626 00:27:54,890 --> 00:27:57,109 Right. And so depending on 627 00:27:57,110 --> 00:27:59,209 how that data is shown, you have 628 00:27:59,210 --> 00:28:01,609 different levels of ability to, 629 00:28:01,610 --> 00:28:03,559 you know, trust it. 630 00:28:03,560 --> 00:28:05,629 Right. Like, I can't I can't actually 631 00:28:05,630 --> 00:28:07,249 go in and look and verify that Google is 632 00:28:07,250 --> 00:28:08,629 telling me the truth here. 633 00:28:08,630 --> 00:28:10,189 I have a better ability to do that from 634 00:28:10,190 --> 00:28:11,190 Tor. 635 00:28:13,270 --> 00:28:15,669 Beyond this, one of the older projects 636 00:28:15,670 --> 00:28:17,739 sort of in this space is crowd 637 00:28:17,740 --> 00:28:19,209 sourcing verdict. 638 00:28:19,210 --> 00:28:20,739 Webb came out of Berkman Center a while 639 00:28:20,740 --> 00:28:22,839 ago and basically asked users 640 00:28:22,840 --> 00:28:24,999 who visit it, shows them 641 00:28:25,000 --> 00:28:26,319 a website and says, can you or can you 642 00:28:26,320 --> 00:28:27,339 not access the site? 643 00:28:27,340 --> 00:28:28,989 And they they check yes or no. 644 00:28:30,210 --> 00:28:32,339 And that, you know, gives you a somewhat 645 00:28:32,340 --> 00:28:34,349 noisy because now you have to trust that 646 00:28:34,350 --> 00:28:36,479 the users are, you know, you 647 00:28:36,480 --> 00:28:38,339 know, is it you know, their connection? 648 00:28:38,340 --> 00:28:39,629 Is that something else? 649 00:28:39,630 --> 00:28:42,059 Like you don't get the same insight 650 00:28:42,060 --> 00:28:43,979 into why something didn't work as you 651 00:28:43,980 --> 00:28:46,249 would if you were running test code? 652 00:28:47,280 --> 00:28:48,280 But you can learn a lot. 653 00:28:49,320 --> 00:28:51,629 And then focusing on specific 654 00:28:51,630 --> 00:28:53,039 countries, you've got things like grade 655 00:28:53,040 --> 00:28:55,499 fire, which basically 656 00:28:55,500 --> 00:28:56,999 runs its own infrastructure in China 657 00:28:57,000 --> 00:28:59,099 specifically, and then keeps monitoring 658 00:28:59,100 --> 00:29:01,379 or sort of keeps tabs on popular 659 00:29:01,380 --> 00:29:02,939 sites that are accessible. 660 00:29:02,940 --> 00:29:03,940 Are not they're. 661 00:29:06,940 --> 00:29:09,219 Switching over to mobile interference, 662 00:29:09,220 --> 00:29:11,559 right, we're entering a world where 663 00:29:11,560 --> 00:29:14,439 cell phones are gaining importance 664 00:29:14,440 --> 00:29:16,719 and we've sort of found ourselves, 665 00:29:16,720 --> 00:29:18,369 you know, having a harder time to 666 00:29:18,370 --> 00:29:19,689 understand exactly what it means to 667 00:29:19,690 --> 00:29:20,589 interfere there. 668 00:29:20,590 --> 00:29:22,899 There's a few projects these range from 669 00:29:22,900 --> 00:29:23,919 sort of Netlist. 670 00:29:23,920 --> 00:29:25,149 It's been around for a while. It's just 671 00:29:25,150 --> 00:29:27,369 sort of a connection analysis or 672 00:29:27,370 --> 00:29:28,629 diagnostics tool. 673 00:29:28,630 --> 00:29:30,909 It lets you see things like, is 674 00:29:30,910 --> 00:29:32,799 my mobile network preventing me from 675 00:29:32,800 --> 00:29:34,509 opening connections on specific ports? 676 00:29:36,040 --> 00:29:38,349 And is there, you know, noticeable 677 00:29:38,350 --> 00:29:39,519 filtering or proxy's 678 00:29:40,960 --> 00:29:43,569 one of the projects that came out in 2015 679 00:29:43,570 --> 00:29:45,909 is something called Asia Chat's. 680 00:29:45,910 --> 00:29:47,739 And that was, again, out of Citizen Lab. 681 00:29:47,740 --> 00:29:49,839 They basically looked into a 682 00:29:49,840 --> 00:29:52,239 bunch of the popular chat networks. 683 00:29:52,240 --> 00:29:54,549 We chat Kakao talk line, looked 684 00:29:54,550 --> 00:29:56,169 for built-In censorship lists. 685 00:29:56,170 --> 00:29:57,639 So is there are there words that you 686 00:29:57,640 --> 00:29:59,919 can't type over these things and 687 00:29:59,920 --> 00:30:01,419 looked at other management policies that 688 00:30:01,420 --> 00:30:03,589 were going on in those specific apps? 689 00:30:03,590 --> 00:30:04,779 Right. We're moving from a world where 690 00:30:04,780 --> 00:30:07,179 what we care about is the open web 691 00:30:07,180 --> 00:30:09,549 and, you know, a domain to 692 00:30:09,550 --> 00:30:11,019 what is happening within a specific 693 00:30:11,020 --> 00:30:12,190 closed app ecosystem. 694 00:30:14,500 --> 00:30:16,579 Sort of the equivalent from 695 00:30:16,580 --> 00:30:18,639 UNNY as we move into mobile is the 696 00:30:18,640 --> 00:30:21,039 library and corresponding 697 00:30:21,040 --> 00:30:23,349 iPhone and Android app around 698 00:30:23,350 --> 00:30:24,939 this project called Measurement Kit, 699 00:30:24,940 --> 00:30:26,409 which is trying to reproduce some of 700 00:30:26,410 --> 00:30:28,539 these connectivity tests, looking at 701 00:30:28,540 --> 00:30:30,729 DNS and HGP that 702 00:30:30,730 --> 00:30:32,859 we that we have run from uni, 703 00:30:32,860 --> 00:30:34,359 but now in a mobile environment. 704 00:30:36,630 --> 00:30:38,399 The other thing that's gaining popularity 705 00:30:38,400 --> 00:30:40,229 is trying to think about what we can 706 00:30:40,230 --> 00:30:42,539 measure externally, so as we've 707 00:30:42,540 --> 00:30:44,759 been sort of, you know, 708 00:30:44,760 --> 00:30:46,559 struggling with this question about 709 00:30:46,560 --> 00:30:48,629 putting things in networks and what risks 710 00:30:48,630 --> 00:30:50,609 where we're sort of exposing people to. 711 00:30:51,960 --> 00:30:53,939 There was a paper and project that came 712 00:30:53,940 --> 00:30:55,139 out sort of at the end of last year, 713 00:30:55,140 --> 00:30:57,509 beginning of this year called Encore, 714 00:30:57,510 --> 00:30:59,639 which brings in the idea of I can 715 00:30:59,640 --> 00:31:01,739 embed the icon, the little image 716 00:31:01,740 --> 00:31:03,689 for Twitter or for Facebook in my page 717 00:31:03,690 --> 00:31:05,819 and see if it loads for my visitors. 718 00:31:05,820 --> 00:31:07,379 And so if all of my visitors from a 719 00:31:07,380 --> 00:31:08,849 specific country fail to load the 720 00:31:08,850 --> 00:31:10,169 Twitter, Facebook, maybe that means 721 00:31:10,170 --> 00:31:11,889 Twitter's blocked. 722 00:31:11,890 --> 00:31:14,039 And so we can have these visitors 723 00:31:14,040 --> 00:31:15,389 who aren't actually cooperating. 724 00:31:15,390 --> 00:31:16,889 They're not running a probe or a 725 00:31:16,890 --> 00:31:18,239 measurement, but we can still learn 726 00:31:18,240 --> 00:31:19,240 things from them. 727 00:31:20,130 --> 00:31:22,199 Similar to that, we have side 728 00:31:22,200 --> 00:31:24,419 channel work and I can and 729 00:31:24,420 --> 00:31:26,319 this guy used for some of the the Tor 730 00:31:26,320 --> 00:31:28,379 accessability stuff as well, which is 731 00:31:28,380 --> 00:31:30,689 saying even when there's two remote 732 00:31:30,690 --> 00:31:32,129 servers that I don't have control of 733 00:31:32,130 --> 00:31:34,469 either of them, I can 734 00:31:34,470 --> 00:31:36,539 using some packet spoofing and TCP 735 00:31:36,540 --> 00:31:38,639 fragmentation leaks, learn if 736 00:31:38,640 --> 00:31:40,949 they can connect to each other or not. 737 00:31:40,950 --> 00:31:43,799 And so now without, you know, putting 738 00:31:43,800 --> 00:31:45,269 those users aren't like cooperating. 739 00:31:45,270 --> 00:31:46,499 They're not running some probing software. 740 00:31:46,500 --> 00:31:48,239 But I as a remote observer, can 741 00:31:48,240 --> 00:31:50,339 potentially learn about connectivity, 742 00:31:50,340 --> 00:31:51,510 your IP level blocks. 743 00:31:54,110 --> 00:31:56,269 We're also still coming to terms with 744 00:31:56,270 --> 00:31:58,240 a world where the IPV for space is small, 745 00:31:59,660 --> 00:32:02,119 so you've got GEMAP, which is 746 00:32:02,120 --> 00:32:04,399 there's a talk on male security 747 00:32:04,400 --> 00:32:06,499 stuff that Zacchia will give 748 00:32:06,500 --> 00:32:08,239 three to three. 749 00:32:08,240 --> 00:32:09,529 But they've also released something 750 00:32:09,530 --> 00:32:11,629 called Sensis just recently, which 751 00:32:11,630 --> 00:32:13,189 is sort of looking at what's running in 752 00:32:13,190 --> 00:32:14,869 different places. 753 00:32:14,870 --> 00:32:16,219 One of the projects that I'm involved 754 00:32:16,220 --> 00:32:18,409 with is called Satellite, which 755 00:32:18,410 --> 00:32:20,479 is looking at open DNS resolvers 756 00:32:20,480 --> 00:32:22,489 and what sort of DNS consistency we can 757 00:32:22,490 --> 00:32:24,559 learn from a single external 758 00:32:24,560 --> 00:32:25,549 vantage point. 759 00:32:25,550 --> 00:32:27,859 So we find roughly eight million 760 00:32:27,860 --> 00:32:30,349 open DNS resolvers that are willing to 761 00:32:30,350 --> 00:32:32,599 resolve domains for you. 762 00:32:32,600 --> 00:32:34,159 And so you ask all of them for popular 763 00:32:34,160 --> 00:32:35,809 domains and you see if there is specific 764 00:32:35,810 --> 00:32:37,969 ISPs or specific regions where you get 765 00:32:37,970 --> 00:32:39,709 weird responses back. 766 00:32:39,710 --> 00:32:41,689 And so now without having, you know, 767 00:32:41,690 --> 00:32:43,249 anything except for these hopefully 768 00:32:43,250 --> 00:32:45,559 infrastructural DNS resolvers, 769 00:32:45,560 --> 00:32:47,239 I can understand if there's some sort of 770 00:32:47,240 --> 00:32:49,489 country or ISP level policy going 771 00:32:49,490 --> 00:32:50,490 on. 772 00:32:52,750 --> 00:32:56,019 All right, so 773 00:32:56,020 --> 00:32:58,179 we're going to sort of recover 774 00:32:58,180 --> 00:32:59,230 about governance, 775 00:33:00,520 --> 00:33:02,649 what's happened actually lately 776 00:33:02,650 --> 00:33:05,269 from the HGP Working Group. 777 00:33:05,270 --> 00:33:07,329 There have been working and it seems that 778 00:33:07,330 --> 00:33:09,429 it's going to be accepted. 779 00:33:09,430 --> 00:33:11,649 The HTP status code four 780 00:33:11,650 --> 00:33:12,650 or five one. 781 00:33:13,870 --> 00:33:15,699 That's a status quo that has that is 782 00:33:15,700 --> 00:33:17,859 going to be used in case 783 00:33:17,860 --> 00:33:20,259 of legal 784 00:33:20,260 --> 00:33:22,419 websites that have been filtered, blocked 785 00:33:22,420 --> 00:33:24,519 or censored because of legal issues or 786 00:33:24,520 --> 00:33:26,079 political incidents. 787 00:33:28,120 --> 00:33:29,979 I hope that this will not be passed. 788 00:33:29,980 --> 00:33:32,169 And this is 789 00:33:32,170 --> 00:33:34,689 somehow like acknowledging the problem 790 00:33:34,690 --> 00:33:36,789 of and saying that's 791 00:33:36,790 --> 00:33:39,189 OK with censor and block 792 00:33:39,190 --> 00:33:40,389 Internet. 793 00:33:40,390 --> 00:33:41,390 But I would rather 794 00:33:43,300 --> 00:33:45,890 protest for a more free Internet. 795 00:33:46,930 --> 00:33:49,119 What did happen to good on 796 00:33:49,120 --> 00:33:51,249 the terms of governments is the 797 00:33:52,510 --> 00:33:55,179 forming of a group called Human Rights 798 00:33:55,180 --> 00:33:57,609 Protocols Consideration Research 799 00:33:57,610 --> 00:33:58,610 Research Group 800 00:33:59,680 --> 00:34:02,049 that is exposing 801 00:34:02,050 --> 00:34:03,999 the relation between protocols and human 802 00:34:04,000 --> 00:34:06,219 rights, having a focus 803 00:34:06,220 --> 00:34:08,319 on the rights of freedom 804 00:34:08,320 --> 00:34:10,899 of expression and freedom of assembly. 805 00:34:12,370 --> 00:34:14,888 And now it's going to 806 00:34:14,889 --> 00:34:15,189 call. 807 00:34:15,190 --> 00:34:17,468 So and I guess the 808 00:34:17,469 --> 00:34:20,229 point of this HRP 809 00:34:20,230 --> 00:34:22,060 group is that they're hoping to 810 00:34:23,380 --> 00:34:26,049 sort of upend the policy on further 811 00:34:26,050 --> 00:34:29,259 IETF standards to include a section 812 00:34:29,260 --> 00:34:31,419 sort of expressing what's going 813 00:34:31,420 --> 00:34:32,948 on in terms of the implications to 814 00:34:32,949 --> 00:34:34,359 freedom of expression and freedom of 815 00:34:34,360 --> 00:34:35,979 assembly with new protocols as they come 816 00:34:35,980 --> 00:34:37,089 in. So this is something that we should 817 00:34:37,090 --> 00:34:38,859 be thinking about as we start 818 00:34:38,860 --> 00:34:39,849 standardizing protocols. 819 00:34:39,850 --> 00:34:40,899 What are the implications? 820 00:34:40,900 --> 00:34:42,759 Are we doing this in a way that is going 821 00:34:42,760 --> 00:34:44,799 to be supporting these universal human 822 00:34:44,800 --> 00:34:45,800 rights? 823 00:34:46,420 --> 00:34:48,519 The other major thing in the 824 00:34:48,520 --> 00:34:50,829 measurement community space is that open 825 00:34:50,830 --> 00:34:53,468 tech fund, a branch of the US government 826 00:34:53,469 --> 00:34:55,569 under a few layers, got 827 00:34:55,570 --> 00:34:57,189 a big injection of like twenty five 828 00:34:57,190 --> 00:34:59,739 billion dollars to support sort of 829 00:34:59,740 --> 00:35:01,449 circumvention and censorship measurement 830 00:35:01,450 --> 00:35:03,639 stuff. And so a bunch of projects 831 00:35:03,640 --> 00:35:05,769 have gotten money over the 832 00:35:05,770 --> 00:35:07,929 last year and that's continuing now. 833 00:35:07,930 --> 00:35:09,579 And so there's been this injection of 834 00:35:09,580 --> 00:35:11,169 sort of excitement and people working on 835 00:35:11,170 --> 00:35:12,399 it as a result. 836 00:35:12,400 --> 00:35:13,749 And so it'll be interesting to see how 837 00:35:13,750 --> 00:35:14,709 long that continues. 838 00:35:14,710 --> 00:35:15,849 But that's really sort of. 839 00:35:16,890 --> 00:35:19,169 Boosted a lot of stuff, but again, 840 00:35:19,170 --> 00:35:20,519 the community is sort of struggling at 841 00:35:20,520 --> 00:35:21,869 the same time with trying to understand 842 00:35:21,870 --> 00:35:23,819 what it means that so much of this work 843 00:35:23,820 --> 00:35:25,739 is now funded by by the U.S.. 844 00:35:28,330 --> 00:35:29,330 OK, 845 00:35:30,880 --> 00:35:32,979 so it's not all roses, although I think 846 00:35:32,980 --> 00:35:35,829 you've already come to to notice that 847 00:35:35,830 --> 00:35:37,419 there's a lot of stuff that the tools 848 00:35:37,420 --> 00:35:38,499 that are out there aren't doing a good 849 00:35:38,500 --> 00:35:39,669 job of measuring. 850 00:35:39,670 --> 00:35:42,159 Just to give you some examples, 851 00:35:42,160 --> 00:35:44,199 in addition to blocking. 852 00:35:44,200 --> 00:35:46,209 Right. Which is a thing that's pretty 853 00:35:46,210 --> 00:35:47,979 binary. I can see if I get the wrong 854 00:35:47,980 --> 00:35:49,779 result or no result. 855 00:35:49,780 --> 00:35:51,159 What happens if I just get the result 856 00:35:51,160 --> 00:35:52,779 really slowly? 857 00:35:52,780 --> 00:35:54,939 My my automatic test is going to have 858 00:35:54,940 --> 00:35:57,219 a hard time coming up with those 859 00:35:57,220 --> 00:35:58,299 heuristics. 860 00:35:58,300 --> 00:35:59,979 And we really haven't written these tests 861 00:35:59,980 --> 00:36:01,509 that actually are doing a good job with 862 00:36:01,510 --> 00:36:03,699 rustics to understand that this is just 863 00:36:03,700 --> 00:36:05,229 throttled and isn't actually going to be 864 00:36:05,230 --> 00:36:07,659 usable and differentiate 865 00:36:07,660 --> 00:36:09,849 that that accessible but really slow 866 00:36:09,850 --> 00:36:10,850 from normal. 867 00:36:11,800 --> 00:36:13,599 We still are pretty far behind on 868 00:36:13,600 --> 00:36:14,769 platform self-censorship. 869 00:36:14,770 --> 00:36:16,269 So our ability to understand what sort of 870 00:36:16,270 --> 00:36:18,639 management and and, 871 00:36:18,640 --> 00:36:20,859 you know, removal of content platforms 872 00:36:20,860 --> 00:36:22,869 are doing by themselves and unilaterally, 873 00:36:24,160 --> 00:36:26,139 like the Greece example points out, a lot 874 00:36:26,140 --> 00:36:28,989 of times policies are not implemented 875 00:36:28,990 --> 00:36:31,059 on a country level, which is 876 00:36:31,060 --> 00:36:33,159 sort of how we try to think about it and 877 00:36:33,160 --> 00:36:34,539 how those maps that we've been showing 878 00:36:34,540 --> 00:36:36,699 show it, but really are at Espy's 879 00:36:36,700 --> 00:36:37,869 or specific regions. 880 00:36:37,870 --> 00:36:39,459 And so we need to tear this data apart 881 00:36:39,460 --> 00:36:42,579 and get down to the specific, you know, 882 00:36:42,580 --> 00:36:44,049 how is it going to work for these home 883 00:36:44,050 --> 00:36:46,059 users in this place? 884 00:36:46,060 --> 00:36:47,229 And it's not a question that we have a 885 00:36:47,230 --> 00:36:48,879 good way to answer right now. 886 00:36:48,880 --> 00:36:50,739 And we also don't have a good handle on 887 00:36:50,740 --> 00:36:51,699 the social impacts. 888 00:36:51,700 --> 00:36:52,629 Do people notice this? 889 00:36:52,630 --> 00:36:55,119 Is this what does it mean to be extreme? 890 00:36:55,120 --> 00:36:57,879 How often are people hitting block pages? 891 00:36:57,880 --> 00:36:59,889 Like is is there other stuff that people 892 00:36:59,890 --> 00:37:01,989 are choosing not to visit, 893 00:37:01,990 --> 00:37:03,639 even though it's not explicitly blocked 894 00:37:03,640 --> 00:37:05,859 because they they 895 00:37:05,860 --> 00:37:07,719 feel unsafe on the Internet because of 896 00:37:07,720 --> 00:37:09,639 the policy like these are these are 897 00:37:09,640 --> 00:37:11,349 things that we really haven't even begun 898 00:37:11,350 --> 00:37:12,350 to explore. 899 00:37:13,880 --> 00:37:16,069 On the academic side, there's been a big 900 00:37:16,070 --> 00:37:18,559 conversation about ethics that's 901 00:37:18,560 --> 00:37:20,509 continued to sort of emerge in the last 902 00:37:20,510 --> 00:37:21,510 year. 903 00:37:22,340 --> 00:37:24,439 Part of that is trying to understand 904 00:37:24,440 --> 00:37:26,569 when it's OK to make measurements right. 905 00:37:26,570 --> 00:37:28,489 If I get abuse complaints, telling me to 906 00:37:28,490 --> 00:37:30,230 knock it off, like 907 00:37:31,730 --> 00:37:33,109 if it's if it's someone who owns an 908 00:37:33,110 --> 00:37:34,639 organization. 909 00:37:34,640 --> 00:37:36,289 Sure. That seems like I probably 910 00:37:36,290 --> 00:37:37,339 shouldn't be doing something they don't 911 00:37:37,340 --> 00:37:38,899 want me to do. What happens if it's a 912 00:37:38,900 --> 00:37:40,369 country that tells me to knock it off? 913 00:37:40,370 --> 00:37:42,169 Do I just not measure their country? 914 00:37:42,170 --> 00:37:44,329 So so where where are these sorts 915 00:37:44,330 --> 00:37:46,129 of things where these boundaries 916 00:37:47,510 --> 00:37:49,669 for our more traditional things inside of 917 00:37:49,670 --> 00:37:50,659 countries? 918 00:37:50,660 --> 00:37:51,859 What does it mean to get informed 919 00:37:51,860 --> 00:37:52,819 consent? 920 00:37:52,820 --> 00:37:54,289 Right. Are the users who are running 921 00:37:54,290 --> 00:37:56,179 probes or who are taking measurements on 922 00:37:56,180 --> 00:37:58,519 these systems, understanding what 923 00:37:58,520 --> 00:38:00,619 they're getting themselves into? 924 00:38:00,620 --> 00:38:02,209 Do we as researchers even know what they 925 00:38:02,210 --> 00:38:03,649 were getting into? And are we able to 926 00:38:03,650 --> 00:38:05,809 tell them those risks in a way 927 00:38:05,810 --> 00:38:07,130 that they can give informed consent? 928 00:38:08,360 --> 00:38:10,399 There haven't been a huge number of, you 929 00:38:10,400 --> 00:38:12,139 know, retributions or anyone getting in 930 00:38:12,140 --> 00:38:13,759 trouble, but we don't know what could 931 00:38:13,760 --> 00:38:15,589 happen. We don't know how technically 932 00:38:15,590 --> 00:38:16,759 savvy the people who would be getting 933 00:38:16,760 --> 00:38:17,659 them in trouble are. And if they 934 00:38:17,660 --> 00:38:19,609 understand the nuances, we don't 935 00:38:19,610 --> 00:38:20,630 understand the nuances. 936 00:38:22,590 --> 00:38:25,109 So these conversations are happening, 937 00:38:25,110 --> 00:38:26,110 a lot of this 938 00:38:28,050 --> 00:38:30,149 there's a bunch of ethicists, weirdly, 939 00:38:30,150 --> 00:38:31,589 it's the US Department of Homeland 940 00:38:31,590 --> 00:38:33,959 Security that that published 941 00:38:33,960 --> 00:38:35,579 something called the Mendelow Report on 942 00:38:35,580 --> 00:38:37,949 Ethical Principles for Communication 943 00:38:37,950 --> 00:38:39,539 Technologies in 2012. 944 00:38:40,980 --> 00:38:43,169 And what that document says 945 00:38:43,170 --> 00:38:44,729 to some extent is that if you can't get 946 00:38:44,730 --> 00:38:46,379 informed consent, you should be 947 00:38:46,380 --> 00:38:48,539 practicing sort of 948 00:38:48,540 --> 00:38:49,589 minimizing risk. 949 00:38:49,590 --> 00:38:50,699 You shouldn't be doing anything 950 00:38:50,700 --> 00:38:51,929 unnecessary in order to get the 951 00:38:51,930 --> 00:38:53,669 measurements that you need. 952 00:38:53,670 --> 00:38:54,959 So now what is at risk? 953 00:38:54,960 --> 00:38:57,119 How do we you know, what measurements do 954 00:38:57,120 --> 00:38:57,599 we need? 955 00:38:57,600 --> 00:38:58,719 Where are these boundaries? 956 00:38:58,720 --> 00:39:00,269 We're still drawing these. 957 00:39:00,270 --> 00:39:02,039 There aren't clear lines yet on any of 958 00:39:02,040 --> 00:39:03,040 this stuff. 959 00:39:03,840 --> 00:39:06,839 But we had the first workshop on ethics 960 00:39:06,840 --> 00:39:08,999 in August at the sitcom 961 00:39:09,000 --> 00:39:11,219 conference. So this is sort of, 962 00:39:11,220 --> 00:39:13,289 I think, at the forefront of a lot 963 00:39:13,290 --> 00:39:15,389 of people's minds in the space is what is 964 00:39:15,390 --> 00:39:17,069 OK, what is safe. 965 00:39:17,070 --> 00:39:19,319 You know, are we are we doing a good job 966 00:39:19,320 --> 00:39:21,389 of both collecting the 967 00:39:21,390 --> 00:39:23,010 measurement data that we need to, 968 00:39:24,300 --> 00:39:25,829 but also doing it in a good way? 969 00:39:27,450 --> 00:39:29,549 So we're going to end 970 00:39:29,550 --> 00:39:31,619 by, you know, the pitch of what you 971 00:39:31,620 --> 00:39:33,359 can do, how you can help. 972 00:39:33,360 --> 00:39:35,519 You can measure, right? 973 00:39:35,520 --> 00:39:37,139 You can collect the data. 974 00:39:37,140 --> 00:39:38,249 You can work on the code. 975 00:39:38,250 --> 00:39:39,869 A lot of these projects are open source. 976 00:39:39,870 --> 00:39:42,119 We'll put these slides on the linked from 977 00:39:42,120 --> 00:39:44,129 the talks so that you can click on all of 978 00:39:44,130 --> 00:39:45,419 them and find the project that you're 979 00:39:45,420 --> 00:39:46,769 most excited about. 980 00:39:46,770 --> 00:39:48,809 You can skip the whole measurement thing 981 00:39:48,810 --> 00:39:50,279 and work directly on circumvention. 982 00:39:50,280 --> 00:39:52,319 Pluggable transports are, you know, 983 00:39:52,320 --> 00:39:54,239 gaining standardization. 984 00:39:54,240 --> 00:39:55,499 There's a nice website, pluggable 985 00:39:55,500 --> 00:39:57,689 transports, dot info that talks about 986 00:39:57,690 --> 00:39:58,919 the spec, which is pretty easy to 987 00:39:58,920 --> 00:40:00,599 implement. And so you can make your own 988 00:40:00,600 --> 00:40:02,729 different way of connecting 989 00:40:02,730 --> 00:40:04,649 and you can have the conversation in your 990 00:40:04,650 --> 00:40:06,359 local jurisdiction or on the Internet as 991 00:40:06,360 --> 00:40:08,459 a whole to advocate for 992 00:40:08,460 --> 00:40:10,559 transparency, if nothing else, so 993 00:40:10,560 --> 00:40:12,539 that you can, you know, understand what 994 00:40:12,540 --> 00:40:14,819 is blocked and figure out where 995 00:40:14,820 --> 00:40:16,169 your community is comfortable having 996 00:40:16,170 --> 00:40:17,170 that. 997 00:40:17,430 --> 00:40:19,949 So we'll be around to answer questions. 998 00:40:19,950 --> 00:40:20,939 Thank you. 999 00:40:20,940 --> 00:40:21,940 Thank you. 1000 00:40:31,120 --> 00:40:33,309 So I assume there'll be some 1001 00:40:33,310 --> 00:40:35,469 questions, OK, 1002 00:40:35,470 --> 00:40:37,119 for questions, please line up. 1003 00:40:37,120 --> 00:40:39,189 There is one, two, three, four mikes here 1004 00:40:39,190 --> 00:40:40,899 and we'll also take some questions from 1005 00:40:40,900 --> 00:40:41,900 the Internet. 1006 00:40:45,590 --> 00:40:47,519 The first gentleman is lined up. 1007 00:40:50,790 --> 00:40:51,779 Go ahead and talk. 1008 00:40:51,780 --> 00:40:53,069 Please talk to the microphone. 1009 00:40:53,070 --> 00:40:54,509 So we got it on tape. 1010 00:40:54,510 --> 00:40:55,859 Thanks. 1011 00:40:55,860 --> 00:40:58,589 As you're probably aware, more and more 1012 00:40:58,590 --> 00:41:00,809 countries, which we do not usually 1013 00:41:00,810 --> 00:41:02,909 think of as restrictive, 1014 00:41:02,910 --> 00:41:05,129 are now blocking something. 1015 00:41:05,130 --> 00:41:07,289 It's not only I suppose it's not 1016 00:41:07,290 --> 00:41:08,999 only because of the attacks. 1017 00:41:09,000 --> 00:41:11,159 So I'd like to I ask for 1018 00:41:11,160 --> 00:41:13,619 advice on advocacy. 1019 00:41:13,620 --> 00:41:15,930 So in my country, there has recently 1020 00:41:16,950 --> 00:41:19,109 been some blocks implemented 1021 00:41:19,110 --> 00:41:20,669 because of lobbyists 1022 00:41:22,230 --> 00:41:24,960 trying to stop you or your 1023 00:41:26,820 --> 00:41:28,949 companies if you don't want to 1024 00:41:28,950 --> 00:41:31,109 go ahead. So and I've been 1025 00:41:31,110 --> 00:41:33,479 trying on behalf of the civil movement to 1026 00:41:33,480 --> 00:41:35,609 advocate to not do that. 1027 00:41:35,610 --> 00:41:38,309 And what I get in return is, well, we 1028 00:41:38,310 --> 00:41:40,619 here's the thing. We need to block this. 1029 00:41:40,620 --> 00:41:42,839 So they are if I give them something 1030 00:41:42,840 --> 00:41:44,429 that can minimize collateral collateral 1031 00:41:44,430 --> 00:41:46,139 damage, they're OK. 1032 00:41:46,140 --> 00:41:48,539 But in my mind, that indicates 1033 00:41:48,540 --> 00:41:49,709 deep packet inspection. 1034 00:41:49,710 --> 00:41:51,989 So is there a recommended 1035 00:41:51,990 --> 00:41:54,119 way, if we are OK to 1036 00:41:54,120 --> 00:41:55,349 block this small thing, but we don't want 1037 00:41:55,350 --> 00:41:58,019 to blow the whole IP, the whole domain, 1038 00:41:58,020 --> 00:42:00,149 do we go DP or is this is there some 1039 00:42:00,150 --> 00:42:01,150 magic pill? 1040 00:42:02,670 --> 00:42:04,109 I don't think there's any magic pills, 1041 00:42:04,110 --> 00:42:05,110 unfortunately. 1042 00:42:06,010 --> 00:42:07,109 I mean, so there's a few things. 1043 00:42:07,110 --> 00:42:09,299 One is whenever you get this sort 1044 00:42:09,300 --> 00:42:11,279 of thing, you need to focus first on 1045 00:42:11,280 --> 00:42:13,319 transparency, because that's something 1046 00:42:13,320 --> 00:42:15,479 that's much easier to get public 1047 00:42:15,480 --> 00:42:16,589 opinion. Right. 1048 00:42:16,590 --> 00:42:18,179 There's always going to be debate about 1049 00:42:18,180 --> 00:42:19,979 where the line of what is acceptable and 1050 00:42:19,980 --> 00:42:21,149 unacceptable content is. 1051 00:42:21,150 --> 00:42:23,279 But I think most people are going to 1052 00:42:23,280 --> 00:42:25,169 be pretty happy to say that we need to 1053 00:42:25,170 --> 00:42:27,389 have public record and public 1054 00:42:27,390 --> 00:42:28,769 knowledge of what is blocked so that we 1055 00:42:28,770 --> 00:42:30,359 can have the conversation, that we can 1056 00:42:30,360 --> 00:42:31,859 audit, that what's being blocked is what 1057 00:42:31,860 --> 00:42:33,899 we've decided is worth blocking. 1058 00:42:33,900 --> 00:42:35,549 So the first thing is if there's these 1059 00:42:35,550 --> 00:42:36,689 regulations or things that are coming 1060 00:42:36,690 --> 00:42:38,699 without that ability for people to 1061 00:42:38,700 --> 00:42:40,589 inspect and understand what's happening, 1062 00:42:40,590 --> 00:42:41,909 that's something that's really important 1063 00:42:41,910 --> 00:42:43,949 and I think much less controversial that 1064 00:42:43,950 --> 00:42:45,029 you can push for. 1065 00:42:45,030 --> 00:42:46,080 And that's worth having 1066 00:42:47,610 --> 00:42:49,139 in terms of should you be blocking 1067 00:42:49,140 --> 00:42:51,209 Phillips and the collateral damage thing. 1068 00:42:54,810 --> 00:42:55,889 That's a hard question. 1069 00:42:55,890 --> 00:42:57,839 It depends a lot. I mean, I think we're 1070 00:42:57,840 --> 00:43:00,629 going to we're going to enter 1071 00:43:00,630 --> 00:43:02,609 and we are we are moving towards a world 1072 00:43:02,610 --> 00:43:04,139 with more encryption through let's 1073 00:43:04,140 --> 00:43:05,519 encrypt in some of these things. 1074 00:43:05,520 --> 00:43:07,679 And so what you could imagine 1075 00:43:07,680 --> 00:43:09,149 is if you're doing things like just 1076 00:43:09,150 --> 00:43:11,759 taking Sanai or specific domains 1077 00:43:11,760 --> 00:43:13,469 and trying to block those, that's 1078 00:43:13,470 --> 00:43:15,779 something that potentially 1079 00:43:15,780 --> 00:43:18,269 technology is going to surpass 1080 00:43:18,270 --> 00:43:20,519 in the next three, five years. 1081 00:43:20,520 --> 00:43:22,169 Something like that will enter a world 1082 00:43:22,170 --> 00:43:24,239 where that sort of block isn't going 1083 00:43:24,240 --> 00:43:25,349 to be nearly as effective. 1084 00:43:25,350 --> 00:43:26,609 Hopefully, we'll get to a world with 1085 00:43:26,610 --> 00:43:28,949 encrypted DNS and with, you know, 1086 00:43:28,950 --> 00:43:31,169 good encryption in 1087 00:43:31,170 --> 00:43:33,719 some of these newer HGP like protocols 1088 00:43:33,720 --> 00:43:36,089 where you can't do these fixes. 1089 00:43:36,090 --> 00:43:38,699 So putting in boxes like that, 1090 00:43:38,700 --> 00:43:41,039 the hope is they become obsolete faster, 1091 00:43:41,040 --> 00:43:42,749 whereas doing IP blocks, that's something 1092 00:43:42,750 --> 00:43:44,879 that's harder to rectify that damage. 1093 00:43:44,880 --> 00:43:47,010 That's that's my personal opinion. 1094 00:43:48,420 --> 00:43:49,469 OK, thank you. 1095 00:43:49,470 --> 00:43:51,449 Tell me back there, Mike, number five, is 1096 00:43:51,450 --> 00:43:53,069 that a question? 1097 00:43:53,070 --> 00:43:55,199 OK, please. Talking to the mike. 1098 00:43:55,200 --> 00:43:57,360 Thanks for taking my question. 1099 00:43:58,560 --> 00:44:00,119 I'm going to ask about the funding that 1100 00:44:00,120 --> 00:44:01,049 you mentioned. 1101 00:44:01,050 --> 00:44:03,119 So you mentioned that the American 1102 00:44:03,120 --> 00:44:05,609 government funds much, if not 1103 00:44:05,610 --> 00:44:08,129 all or most of 1104 00:44:08,130 --> 00:44:09,779 the work that you do and a lot of the 1105 00:44:09,780 --> 00:44:12,419 work on censorship research. 1106 00:44:13,920 --> 00:44:16,269 I'd like you to talk more about that. 1107 00:44:16,270 --> 00:44:17,579 What do you make of it? 1108 00:44:17,580 --> 00:44:20,219 How do you think this fits? 1109 00:44:20,220 --> 00:44:22,229 Your work fits into America's foreign 1110 00:44:22,230 --> 00:44:24,569 policy objectives and if it does 1111 00:44:24,570 --> 00:44:25,570 in any way. 1112 00:44:26,520 --> 00:44:27,659 Sure. 1113 00:44:27,660 --> 00:44:29,579 So so I guess what I'll say is there are 1114 00:44:29,580 --> 00:44:30,749 other sources of funding to 1115 00:44:31,770 --> 00:44:33,149 a bunch of this stuff happens from 1116 00:44:33,150 --> 00:44:35,309 academics who get funded through through 1117 00:44:35,310 --> 00:44:38,849 grants and through the university 1118 00:44:38,850 --> 00:44:40,949 grant process rather than 1119 00:44:40,950 --> 00:44:42,419 directly through through governmental 1120 00:44:42,420 --> 00:44:43,559 funding. 1121 00:44:43,560 --> 00:44:45,629 I think it's reasonable to see 1122 00:44:45,630 --> 00:44:47,909 this as in line with U.S. 1123 00:44:47,910 --> 00:44:48,989 foreign policy. 1124 00:44:48,990 --> 00:44:50,939 Right. They they would like to have 1125 00:44:52,020 --> 00:44:54,419 generally that 1126 00:44:54,420 --> 00:44:55,409 freedom of expression. 1127 00:44:55,410 --> 00:44:57,239 That's something that, you know, I think 1128 00:44:57,240 --> 00:44:59,009 that that group at least. 1129 00:44:59,010 --> 00:45:01,199 Right, the the Radio Free 1130 00:45:01,200 --> 00:45:03,779 Asia Broadcasting Board of Governors 1131 00:45:03,780 --> 00:45:05,219 is pretty in favor of. 1132 00:45:06,390 --> 00:45:08,339 And so I think for the projects where 1133 00:45:08,340 --> 00:45:10,529 their incentives align, it's 1134 00:45:10,530 --> 00:45:13,049 not like you're not agreeing 1135 00:45:13,050 --> 00:45:14,279 to do anything right. 1136 00:45:14,280 --> 00:45:15,929 What it's potentially doing is it's 1137 00:45:15,930 --> 00:45:17,789 shifting the work. 1138 00:45:17,790 --> 00:45:19,319 Right. And so what we need to be careful 1139 00:45:19,320 --> 00:45:20,489 about is that 1140 00:45:21,660 --> 00:45:23,129 the work that we're doing isn't getting 1141 00:45:23,130 --> 00:45:25,289 shifted purely to circumvention and away 1142 00:45:25,290 --> 00:45:26,969 from something like surveillance, which 1143 00:45:26,970 --> 00:45:29,459 is maybe less exciting or less less 1144 00:45:29,460 --> 00:45:30,359 funded. 1145 00:45:30,360 --> 00:45:31,649 And so that's the thing where we need to 1146 00:45:31,650 --> 00:45:33,809 be careful. But I think that, you know, 1147 00:45:33,810 --> 00:45:35,219 the specific things that they're funding 1148 00:45:35,220 --> 00:45:36,749 are all great projects that I'm happy to 1149 00:45:36,750 --> 00:45:38,729 see getting that injection of support. 1150 00:45:40,590 --> 00:45:42,899 And I think, you know, the call 1151 00:45:42,900 --> 00:45:44,969 that I would make is that we should have 1152 00:45:44,970 --> 00:45:46,889 more people also stepping up to fund. 1153 00:45:46,890 --> 00:45:48,929 I think that's something that you know, 1154 00:45:48,930 --> 00:45:50,999 what I've heard from Open Tech Fund is 1155 00:45:51,000 --> 00:45:52,139 they would also like to see that. 1156 00:45:52,140 --> 00:45:53,369 Right. This isn't something that's 1157 00:45:53,370 --> 00:45:54,959 sustainable from a single government 1158 00:45:54,960 --> 00:45:57,059 because of, you know, the optics don't 1159 00:45:57,060 --> 00:45:59,729 necessarily look great and also because 1160 00:45:59,730 --> 00:46:01,319 why should they be able to make those 1161 00:46:01,320 --> 00:46:02,219 decisions? 1162 00:46:02,220 --> 00:46:03,989 And so the way that we counter that is 1163 00:46:03,990 --> 00:46:06,299 not by refusing to take money, 1164 00:46:06,300 --> 00:46:08,549 but it's by saying, hey, maybe Germany 1165 00:46:08,550 --> 00:46:09,659 should be giving more money, maybe a 1166 00:46:09,660 --> 00:46:11,129 bunch of these other countries that care 1167 00:46:11,130 --> 00:46:13,259 about these same, you know, freedom of 1168 00:46:13,260 --> 00:46:15,209 expression issues should also be stepping 1169 00:46:15,210 --> 00:46:16,210 up to help us. 1170 00:46:17,070 --> 00:46:19,229 What's the mix right now? 1171 00:46:19,230 --> 00:46:21,239 Private to government funding off the top 1172 00:46:21,240 --> 00:46:22,240 of your head? 1173 00:46:22,910 --> 00:46:25,009 That is like I have 1174 00:46:25,010 --> 00:46:26,089 no idea. 1175 00:46:26,090 --> 00:46:27,499 There's a bunch of companies, there's a 1176 00:46:27,500 --> 00:46:29,719 bunch of VPs and stuff that are making 1177 00:46:29,720 --> 00:46:31,789 money from users, they're 1178 00:46:31,790 --> 00:46:33,379 not going to tell you how much money they 1179 00:46:33,380 --> 00:46:35,599 make, but there are pretty big players 1180 00:46:35,600 --> 00:46:36,600 in the space. 1181 00:46:39,480 --> 00:46:41,609 OTF, right, they do publish 1182 00:46:41,610 --> 00:46:42,779 that they're their grant from the 1183 00:46:42,780 --> 00:46:43,979 government. Twenty five million, so we've 1184 00:46:43,980 --> 00:46:45,479 got that. But I don't know what how big 1185 00:46:45,480 --> 00:46:46,710 the overall spaces. 1186 00:46:48,480 --> 00:46:50,549 OK, we 1187 00:46:50,550 --> 00:46:52,709 are a question from the Internet here. 1188 00:46:52,710 --> 00:46:53,819 Yes. 1189 00:46:53,820 --> 00:46:55,619 So talking to your mike, please. 1190 00:46:55,620 --> 00:46:57,299 This is a question from Jay Nixon. 1191 00:46:57,300 --> 00:46:59,429 Hamiet, what's better 1192 00:46:59,430 --> 00:47:01,949 about recognizing censorship if I record 1193 00:47:01,950 --> 00:47:04,469 five one and 1194 00:47:04,470 --> 00:47:06,359 because the page was blocked nonetheless 1195 00:47:06,360 --> 00:47:07,949 and it's good to know what happened. 1196 00:47:07,950 --> 00:47:09,179 Right. 1197 00:47:09,180 --> 00:47:11,429 OK, so I think that by having 1198 00:47:11,430 --> 00:47:13,719 an established code to 1199 00:47:13,720 --> 00:47:15,839 recognize censorship is like Bagnolet, 1200 00:47:15,840 --> 00:47:17,039 as I mentioned, the president will 1201 00:47:17,040 --> 00:47:19,079 acknowledge the fact that, OK, yeah, 1202 00:47:19,080 --> 00:47:20,999 censorship is for granted. 1203 00:47:21,000 --> 00:47:23,249 And we should assign a status code, 1204 00:47:23,250 --> 00:47:25,379 I guess, to my opinion, instead of 1205 00:47:25,380 --> 00:47:27,239 just assigning new status code of 1206 00:47:27,240 --> 00:47:28,320 something that we can't 1207 00:47:29,850 --> 00:47:32,069 really happen, like like 1208 00:47:32,070 --> 00:47:33,929 Internet should not be censored. 1209 00:47:33,930 --> 00:47:36,449 So assigning a status quo to the 1210 00:47:36,450 --> 00:47:37,439 piece would not happen. 1211 00:47:37,440 --> 00:47:39,749 So if if they would like to do it, 1212 00:47:39,750 --> 00:47:40,979 we do it in a different way. 1213 00:47:40,980 --> 00:47:43,139 But assigning like changing 1214 00:47:43,140 --> 00:47:45,090 the status code and assigning and 1215 00:47:46,440 --> 00:47:48,599 acknowledging the fact that we 1216 00:47:48,600 --> 00:47:50,789 still have more, you 1217 00:47:50,790 --> 00:47:52,619 know, like a blocking code. 1218 00:47:52,620 --> 00:47:54,779 No, I, I don't think that is 1219 00:47:54,780 --> 00:47:55,919 a way to go. 1220 00:47:55,920 --> 00:47:58,139 I'm I don't 1221 00:47:58,140 --> 00:47:59,399 have any alternatives on that. 1222 00:47:59,400 --> 00:48:01,969 But, you know, I mean, why 1223 00:48:01,970 --> 00:48:04,049 why we need to to have alternative 1224 00:48:04,050 --> 00:48:06,519 on blogging then they do it otherwise 1225 00:48:06,520 --> 00:48:07,509 without status code. 1226 00:48:07,510 --> 00:48:09,689 So why do give the freedom of using also 1227 00:48:09,690 --> 00:48:11,859 studies because OK, so these websites 1228 00:48:11,860 --> 00:48:12,959 ability and the status quo. 1229 00:48:12,960 --> 00:48:15,059 So now we know that OK, know these 1230 00:48:15,060 --> 00:48:16,379 sites are blocked. Yeah. 1231 00:48:16,380 --> 00:48:17,579 They will still block. 1232 00:48:17,580 --> 00:48:19,289 Still block. 1233 00:48:19,290 --> 00:48:21,509 As I mentioned, the presentation started 1234 00:48:21,510 --> 00:48:24,149 sites and resources legally 1235 00:48:24,150 --> 00:48:25,169 and illegally. 1236 00:48:25,170 --> 00:48:27,089 So now that we have also the four or five 1237 00:48:27,090 --> 00:48:28,860 one stats of the legal. 1238 00:48:30,030 --> 00:48:32,069 So this is a legal like status code. 1239 00:48:32,070 --> 00:48:34,379 And, you know, this will continue. 1240 00:48:34,380 --> 00:48:36,509 They will continue blogging again, you 1241 00:48:36,510 --> 00:48:37,820 know, websites illegally. 1242 00:48:40,590 --> 00:48:41,679 OK, thank you. 1243 00:48:41,680 --> 00:48:44,009 Number two years ago. 1244 00:48:44,010 --> 00:48:44,939 I thank you for it. 1245 00:48:44,940 --> 00:48:47,129 OK, I have a question about talking 1246 00:48:47,130 --> 00:48:47,969 to you. Mike, please. 1247 00:48:47,970 --> 00:48:50,249 I have a question about the gambling, 1248 00:48:51,750 --> 00:48:54,539 about the gambling censorship, 1249 00:48:54,540 --> 00:48:56,129 because this seems like the beginning of 1250 00:48:56,130 --> 00:48:58,349 a general kind of censorship. 1251 00:48:58,350 --> 00:49:00,119 Do you have any idea if anywhere in 1252 00:49:00,120 --> 00:49:02,159 Europe it was possible to get this 1253 00:49:02,160 --> 00:49:04,229 removed in some way by any 1254 00:49:04,230 --> 00:49:06,419 legal measures, lawful measures, 1255 00:49:06,420 --> 00:49:07,979 something like that? 1256 00:49:07,980 --> 00:49:10,089 Well, actually, most of of these 1257 00:49:10,090 --> 00:49:11,729 blogs are quite trivial. 1258 00:49:11,730 --> 00:49:13,889 They are there. 1259 00:49:13,890 --> 00:49:16,289 They were doing the DNS blocking. 1260 00:49:16,290 --> 00:49:18,539 So you could just change your name, 1261 00:49:18,540 --> 00:49:20,729 server, your resolver, and you could look 1262 00:49:20,730 --> 00:49:22,469 out around this issue. 1263 00:49:23,520 --> 00:49:25,709 Many of these sites can be 1264 00:49:25,710 --> 00:49:27,869 seen from other like 1265 00:49:27,870 --> 00:49:30,179 they can be seen by via Tor or 1266 00:49:30,180 --> 00:49:32,489 they they are just trying to 1267 00:49:33,600 --> 00:49:35,339 block these locally. 1268 00:49:35,340 --> 00:49:37,499 They were they were just trying to do 1269 00:49:37,500 --> 00:49:38,579 this in a local way. 1270 00:49:38,580 --> 00:49:40,679 They are not they are not doing 1271 00:49:40,680 --> 00:49:42,059 it well right now. 1272 00:49:42,060 --> 00:49:43,290 They are just 1273 00:49:45,330 --> 00:49:47,699 like to, let's 1274 00:49:47,700 --> 00:49:50,009 say, stop viewing this content from 1275 00:49:50,010 --> 00:49:52,379 the non power users. 1276 00:49:52,380 --> 00:49:53,699 Let's say so. 1277 00:49:55,680 --> 00:49:58,049 So I come from Bulgaria and there they 1278 00:49:58,050 --> 00:50:00,239 switched from just DNS blocking to 1279 00:50:00,240 --> 00:50:02,339 IP blocking, and they've been 1280 00:50:02,340 --> 00:50:03,689 doing that pretty thoroughly for the 1281 00:50:03,690 --> 00:50:06,299 sites they were ordered to block, which 1282 00:50:06,300 --> 00:50:08,039 results in a lot of collateral damage for 1283 00:50:08,040 --> 00:50:10,229 this. And 1284 00:50:10,230 --> 00:50:12,449 we know that, yeah, using tor using 1285 00:50:12,450 --> 00:50:13,799 any kind of it can just go around. 1286 00:50:13,800 --> 00:50:16,169 But it's extremely trivial, but it 1287 00:50:16,170 --> 00:50:17,489 creates a lot of problems. 1288 00:50:17,490 --> 00:50:19,409 It creates a very bad precedent for 1289 00:50:19,410 --> 00:50:21,329 everything else, because right now it's 1290 00:50:21,330 --> 00:50:22,709 gambling sites. Then we have the 1291 00:50:22,710 --> 00:50:25,019 infrastructure and then anything else. 1292 00:50:25,020 --> 00:50:27,209 So are you aware of anyone that 1293 00:50:27,210 --> 00:50:29,729 is working to stop this on a political 1294 00:50:29,730 --> 00:50:30,929 level? 1295 00:50:30,930 --> 00:50:32,429 We should we should speak to these 1296 00:50:32,430 --> 00:50:33,899 people. We should do stuff like 1297 00:50:33,900 --> 00:50:35,419 maintaining that, hey, you are destroying 1298 00:50:35,420 --> 00:50:37,409 the Internet. You know, you can just 1299 00:50:37,410 --> 00:50:38,789 start blocking IP. 1300 00:50:38,790 --> 00:50:40,139 You know that there are like thousands of 1301 00:50:40,140 --> 00:50:42,299 websites hosted. 1302 00:50:42,300 --> 00:50:43,889 Right. So, again, this is this is the 1303 00:50:43,890 --> 00:50:44,699 place for advocacy. 1304 00:50:44,700 --> 00:50:46,949 Right. Finding your local groups that are 1305 00:50:46,950 --> 00:50:48,179 caring about Internet freedom, 1306 00:50:49,470 --> 00:50:51,449 working with them to find especially 1307 00:50:51,450 --> 00:50:52,859 instances of collateral damage are a 1308 00:50:52,860 --> 00:50:55,049 great place for advocacy of look at these 1309 00:50:55,050 --> 00:50:57,119 sites that for whatever reason, are 1310 00:50:57,120 --> 00:50:58,049 being blocked. 1311 00:50:58,050 --> 00:50:59,759 This isn't a cool thing to be doing. 1312 00:50:59,760 --> 00:51:01,229 We really are overstepping. 1313 00:51:01,230 --> 00:51:02,669 We're losing transparency. 1314 00:51:02,670 --> 00:51:04,469 These are all great arguments that you 1315 00:51:04,470 --> 00:51:06,599 can take to 1316 00:51:06,600 --> 00:51:08,879 your politicians and lawmakers and 1317 00:51:08,880 --> 00:51:10,109 and find the ones who are going to help 1318 00:51:10,110 --> 00:51:12,059 champion the issues for you. 1319 00:51:12,060 --> 00:51:14,309 We tried that didn't work. 1320 00:51:14,310 --> 00:51:15,239 Thank you. 1321 00:51:15,240 --> 00:51:17,069 OK, thank you. 1322 00:51:17,070 --> 00:51:18,299 Number six years ago. 1323 00:51:19,440 --> 00:51:22,139 I thank you very much for an interesting 1324 00:51:22,140 --> 00:51:24,359 overview. I wondered if you 1325 00:51:24,360 --> 00:51:26,099 would like to share some thoughts on 1326 00:51:26,100 --> 00:51:28,199 maybe combining or proposals 1327 00:51:28,200 --> 00:51:30,989 to combine censorship measurement 1328 00:51:30,990 --> 00:51:34,439 with censorship circumvention. 1329 00:51:34,440 --> 00:51:36,449 There have been some proposals recently 1330 00:51:36,450 --> 00:51:37,799 for tools that could bind the. 1331 00:51:37,800 --> 00:51:38,909 And I just wondered what your thoughts 1332 00:51:38,910 --> 00:51:39,910 were on that. 1333 00:51:41,010 --> 00:51:42,689 Yeah, so there was there was a there was 1334 00:51:42,690 --> 00:51:45,809 a paper that came out about 1335 00:51:45,810 --> 00:51:47,759 the circumvention tools, our way to 1336 00:51:47,760 --> 00:51:48,760 understand measurement. 1337 00:51:50,340 --> 00:51:52,679 Again, you've got exactly the same 1338 00:51:52,680 --> 00:51:55,049 issues about 1339 00:51:55,050 --> 00:51:57,039 what are the risks that users are facing. 1340 00:51:57,040 --> 00:51:59,909 Right. And and that's now maybe 1341 00:51:59,910 --> 00:52:02,399 you're having a harder argument 1342 00:52:02,400 --> 00:52:04,559 about not 1343 00:52:04,560 --> 00:52:05,969 knowing what's going on, that you know, 1344 00:52:05,970 --> 00:52:08,459 that this is a user who is actively, 1345 00:52:08,460 --> 00:52:09,899 you know, worried about something and 1346 00:52:09,900 --> 00:52:11,249 trying to get around something. 1347 00:52:11,250 --> 00:52:13,349 And so are you now comfortable trying 1348 00:52:13,350 --> 00:52:14,939 those sites that are blocked, which are 1349 00:52:14,940 --> 00:52:16,709 the ones that are being looked for? 1350 00:52:16,710 --> 00:52:18,179 And what sort of dangers are you getting 1351 00:52:18,180 --> 00:52:19,180 your user into? 1352 00:52:20,100 --> 00:52:22,259 At the same time, there are 1353 00:52:22,260 --> 00:52:24,299 plenty of tools that are already doing 1354 00:52:24,300 --> 00:52:26,609 this right. You've got a set of tools 1355 00:52:26,610 --> 00:52:28,409 that do fallback stuff. 1356 00:52:28,410 --> 00:52:29,999 So they first try to make the request 1357 00:52:30,000 --> 00:52:31,949 locally and then if it doesn't work, they 1358 00:52:31,950 --> 00:52:32,999 run it through a proxy. 1359 00:52:33,000 --> 00:52:34,439 So they're already exposing the user to 1360 00:52:34,440 --> 00:52:36,239 that risk. You might as well collect that 1361 00:52:36,240 --> 00:52:37,240 data. Right. 1362 00:52:37,770 --> 00:52:39,959 And so then it's just a measure of 1363 00:52:39,960 --> 00:52:41,729 what sort of anonymity and protection 1364 00:52:41,730 --> 00:52:43,949 against that collected data do you get 1365 00:52:43,950 --> 00:52:46,539 so that if that data gets public because 1366 00:52:46,540 --> 00:52:48,239 expect data to leak to the people, you 1367 00:52:48,240 --> 00:52:49,979 don't want it to, how does it not get 1368 00:52:49,980 --> 00:52:50,999 anyone in trouble? Right. 1369 00:52:51,000 --> 00:52:52,499 Because it's easier to just take that 1370 00:52:52,500 --> 00:52:54,509 whole database of all of the users in the 1371 00:52:54,510 --> 00:52:55,829 country who tried to visit these bad 1372 00:52:55,830 --> 00:52:57,869 sites than having to do the network like 1373 00:52:57,870 --> 00:52:59,099 measurements of them individually, and 1374 00:52:59,100 --> 00:53:01,289 that fortunately, they do collect data of 1375 00:53:01,290 --> 00:53:02,729 these users and they don't know what to 1376 00:53:02,730 --> 00:53:03,719 do with these data. 1377 00:53:03,720 --> 00:53:05,779 Some sense of perceived 1378 00:53:05,780 --> 00:53:07,050 synchronization tools. 1379 00:53:08,550 --> 00:53:09,089 Thank you. 1380 00:53:09,090 --> 00:53:11,249 OK, another question 1381 00:53:11,250 --> 00:53:13,199 from the Internet and please talking to 1382 00:53:13,200 --> 00:53:14,200 much. 1383 00:53:15,030 --> 00:53:16,469 All right. This is a question from 1384 00:53:16,470 --> 00:53:17,470 Jennings. 1385 00:53:18,360 --> 00:53:20,429 It's a two parter. So there are 1386 00:53:20,430 --> 00:53:22,519 a lot of methods for future evasion 1387 00:53:22,520 --> 00:53:24,989 just on those ICMP DNS tunnels. 1388 00:53:24,990 --> 00:53:27,239 That being said, is there is any 1389 00:53:27,240 --> 00:53:29,189 government able to effectively censor a 1390 00:53:29,190 --> 00:53:31,349 block experience users 1391 00:53:31,350 --> 00:53:32,820 who are able to use these, 1392 00:53:33,900 --> 00:53:36,089 I guess, tunnels? 1393 00:53:36,090 --> 00:53:39,239 And is it too paranoid 1394 00:53:39,240 --> 00:53:41,789 to believe that they might intentionally 1395 00:53:41,790 --> 00:53:43,439 allow the government might intentionally 1396 00:53:43,440 --> 00:53:45,539 allow some protocols that they can 1397 00:53:45,540 --> 00:53:46,540 decode 1398 00:53:47,640 --> 00:53:49,769 because, you know, they want 1399 00:53:49,770 --> 00:53:52,019 to actually inspect what people are 1400 00:53:52,020 --> 00:53:53,020 doing. 1401 00:53:54,690 --> 00:53:56,729 It is not too paranoid to expect that 1402 00:53:56,730 --> 00:53:58,949 there are instances that, you know, 1403 00:53:58,950 --> 00:54:01,259 have shown up of proxy's 1404 00:54:01,260 --> 00:54:03,329 or other things that are being run 1405 00:54:03,330 --> 00:54:05,339 by governments to be able to get a sense 1406 00:54:05,340 --> 00:54:06,360 of what users are doing. 1407 00:54:08,340 --> 00:54:10,589 If you are looking for your technically 1408 00:54:10,590 --> 00:54:12,449 competent adversary, you're probably 1409 00:54:12,450 --> 00:54:14,129 looking to China. 1410 00:54:14,130 --> 00:54:16,169 We've seen different things from that 1411 00:54:16,170 --> 00:54:17,549 network. 1412 00:54:17,550 --> 00:54:19,649 And because of how big it 1413 00:54:19,650 --> 00:54:21,119 is and the diversity of ISPs, you see 1414 00:54:21,120 --> 00:54:22,049 different things and you get mixed 1415 00:54:22,050 --> 00:54:23,639 reports. But there's a sense that at 1416 00:54:23,640 --> 00:54:24,779 least some of the time you get 1417 00:54:25,800 --> 00:54:26,800 weirdly 1418 00:54:27,930 --> 00:54:29,759 like long term connections that don't 1419 00:54:29,760 --> 00:54:31,589 look like teals will get throttled or 1420 00:54:31,590 --> 00:54:33,299 become, you know, flaky. 1421 00:54:33,300 --> 00:54:35,339 And that affects a lot of these tunneling 1422 00:54:35,340 --> 00:54:37,649 solutions. We still need more work on 1423 00:54:37,650 --> 00:54:39,809 being able to get reliable things. 1424 00:54:39,810 --> 00:54:42,509 Maybe you use ISP or IP diversity, 1425 00:54:42,510 --> 00:54:43,919 maybe you use lots of short lived 1426 00:54:43,920 --> 00:54:45,809 connections. UDP, there's a bunch of 1427 00:54:45,810 --> 00:54:47,189 different things out there to try. 1428 00:54:47,190 --> 00:54:49,289 There aren't good things that seem 1429 00:54:49,290 --> 00:54:51,269 that, you know, everyone feels are 1430 00:54:51,270 --> 00:54:53,879 reliable and are hard to 1431 00:54:53,880 --> 00:54:55,349 find for a technically competent 1432 00:54:55,350 --> 00:54:56,550 adversary to block. 1433 00:54:58,560 --> 00:55:00,929 OK, there's somebody waiting 1434 00:55:00,930 --> 00:55:02,129 on number five. 1435 00:55:02,130 --> 00:55:03,320 Go ahead talking to them. 1436 00:55:04,500 --> 00:55:05,759 Yes. 1437 00:55:05,760 --> 00:55:07,889 About I think it was 1438 00:55:07,890 --> 00:55:09,959 around six months ago, there was a Web 1439 00:55:09,960 --> 00:55:11,879 site that was forced to shut down that 1440 00:55:11,880 --> 00:55:13,289 archive, political tweets from 1441 00:55:13,290 --> 00:55:14,290 politicians, 1442 00:55:15,750 --> 00:55:17,879 essentially the politicians strong 1443 00:55:17,880 --> 00:55:20,129 arm them into shutting 1444 00:55:20,130 --> 00:55:22,739 down. My question is, 1445 00:55:22,740 --> 00:55:25,049 what kind of 1446 00:55:25,050 --> 00:55:27,479 implementation or capabilities 1447 00:55:27,480 --> 00:55:29,759 would you suggest to try 1448 00:55:29,760 --> 00:55:32,039 to prevent that kind of strongarming 1449 00:55:32,040 --> 00:55:34,529 from happening again with archiving 1450 00:55:34,530 --> 00:55:36,719 suppressed tweets and things 1451 00:55:36,720 --> 00:55:37,859 like that? 1452 00:55:37,860 --> 00:55:39,659 So depending on what threats you are 1453 00:55:39,660 --> 00:55:41,249 worried about, we've got a number of 1454 00:55:41,250 --> 00:55:42,839 technologies that our community knows 1455 00:55:42,840 --> 00:55:44,549 about. We've got hidden services. 1456 00:55:44,550 --> 00:55:46,289 If you want to run something where it's 1457 00:55:46,290 --> 00:55:49,229 IP address and the server is not known. 1458 00:55:49,230 --> 00:55:51,269 So if you feel like the infrastructure 1459 00:55:51,270 --> 00:55:52,559 you know, is in danger, 1460 00:55:53,700 --> 00:55:55,949 there are ways to make that site run in 1461 00:55:55,950 --> 00:55:57,239 a way that it is anonymous. 1462 00:55:58,590 --> 00:56:00,269 You can look at the different legal 1463 00:56:00,270 --> 00:56:02,309 jurisdictions and come with a 1464 00:56:02,310 --> 00:56:04,559 jurisdiction where you are, 1465 00:56:04,560 --> 00:56:06,539 you know, not going to have the same 1466 00:56:06,540 --> 00:56:07,499 ability. Right. 1467 00:56:07,500 --> 00:56:08,489 If it's in a different country, it 1468 00:56:08,490 --> 00:56:09,490 becomes much harder. 1469 00:56:11,400 --> 00:56:13,499 And then you can find a set of 1470 00:56:13,500 --> 00:56:14,999 lawyers. If you believe this is legal, 1471 00:56:15,000 --> 00:56:17,339 which I mean, in a lot of places it is, 1472 00:56:18,480 --> 00:56:19,439 you just need to find the right 1473 00:56:19,440 --> 00:56:20,999 representation where you don't give in 1474 00:56:21,000 --> 00:56:22,559 and show that this actually is an 1475 00:56:22,560 --> 00:56:23,560 instance of free speech. 1476 00:56:25,620 --> 00:56:27,419 If I remember correctly, they actually 1477 00:56:27,420 --> 00:56:28,949 strongarm Twitter into revoking the 1478 00:56:28,950 --> 00:56:31,169 developer access of the website 1479 00:56:31,170 --> 00:56:32,159 in question. 1480 00:56:32,160 --> 00:56:34,289 OK, so then it's things like Free Wabo 1481 00:56:34,290 --> 00:56:36,629 that are crawling and scraping, 1482 00:56:36,630 --> 00:56:38,759 right? We you can't expect that 1483 00:56:38,760 --> 00:56:40,349 the platform is necessarily going to be 1484 00:56:40,350 --> 00:56:41,339 happy about that. 1485 00:56:41,340 --> 00:56:43,019 You can potentially raise a stink and get 1486 00:56:43,020 --> 00:56:45,149 the platforms to reverse those 1487 00:56:45,150 --> 00:56:46,150 sorts of policies. 1488 00:56:47,220 --> 00:56:49,139 Or you can try and show the value of the 1489 00:56:49,140 --> 00:56:50,819 system so that they realize that this is 1490 00:56:50,820 --> 00:56:52,559 something that is worth having. 1491 00:56:52,560 --> 00:56:54,719 Right. In the same way that we have 1492 00:56:54,720 --> 00:56:56,489 things like chilling effects, archiving 1493 00:56:56,490 --> 00:56:58,739 the DMCA takedown request, it 1494 00:56:58,740 --> 00:57:00,509 may be worth having, you know, this sort 1495 00:57:00,510 --> 00:57:02,969 of transparency and trying to convince 1496 00:57:02,970 --> 00:57:04,229 the platforms. 1497 00:57:04,230 --> 00:57:06,239 But at the same time, you can again make 1498 00:57:06,240 --> 00:57:07,799 the argument that even without the 1499 00:57:07,800 --> 00:57:09,479 platform support, there's there's value 1500 00:57:09,480 --> 00:57:10,769 in doing this sort of thing. 1501 00:57:10,770 --> 00:57:11,069 All right. 1502 00:57:11,070 --> 00:57:12,089 Thank you. 1503 00:57:12,090 --> 00:57:13,889 Thank you. We have a tight schedule. 1504 00:57:13,890 --> 00:57:15,599 So this will be the last question were 1505 00:57:15,600 --> 00:57:17,579 taken from before. 1506 00:57:17,580 --> 00:57:19,439 Is there anybody else you'll have to come 1507 00:57:19,440 --> 00:57:20,379 up front for later? 1508 00:57:20,380 --> 00:57:22,079 No, for go ahead. 1509 00:57:22,080 --> 00:57:23,819 Just a short question. 1510 00:57:23,820 --> 00:57:26,129 You mentioned EPP's very shortly. 1511 00:57:26,130 --> 00:57:28,449 There is like this debate, especially 1512 00:57:28,450 --> 00:57:30,989 around Apple, that they have like 1513 00:57:30,990 --> 00:57:33,239 a very special way of thinking about, 1514 00:57:33,240 --> 00:57:34,979 like what is a good app and what not. 1515 00:57:34,980 --> 00:57:37,139 And is there anybody out 1516 00:57:37,140 --> 00:57:39,359 there who measures like taken 1517 00:57:39,360 --> 00:57:40,769 like apps that are not approved or 1518 00:57:40,770 --> 00:57:41,939 something like that? 1519 00:57:41,940 --> 00:57:43,739 Is there something like this? 1520 00:57:43,740 --> 00:57:45,599 Or don't you consider this as censorship 1521 00:57:45,600 --> 00:57:47,189 because it's from a company and not from 1522 00:57:47,190 --> 00:57:48,719 a state that absolutely seems valuable? 1523 00:57:48,720 --> 00:57:49,919 I don't know of any projects that are 1524 00:57:49,920 --> 00:57:51,659 doing a good job like that are that are 1525 00:57:51,660 --> 00:57:52,979 out there yet doing that. 1526 00:57:52,980 --> 00:57:55,339 So I'd love to hear about one. 1527 00:57:55,340 --> 00:57:57,479 And I again, there's been reports of 1528 00:57:57,480 --> 00:57:59,189 the Google Play store entering China 1529 00:57:59,190 --> 00:58:00,929 that's going to involve some sort of 1530 00:58:00,930 --> 00:58:02,969 management of what apps are allowed. 1531 00:58:02,970 --> 00:58:05,069 And again, like keeping track of that 1532 00:58:05,070 --> 00:58:07,019 stuff is maybe maybe someone could do 1533 00:58:07,020 --> 00:58:07,939 that because I think. 1534 00:58:07,940 --> 00:58:08,940 Quite important today. 1535 00:58:10,420 --> 00:58:12,669 I agree, if you want 1536 00:58:12,670 --> 00:58:14,260 to do this, please come and find us. 1537 00:58:16,720 --> 00:58:18,849 OK, we close this talk 1538 00:58:18,850 --> 00:58:20,269 now because we're tight on schedule. 1539 00:58:20,270 --> 00:58:21,969 Thank you very much. Let's have a last 1540 00:58:21,970 --> 00:58:22,970 hand. Thanks.