WEBVTT - Which Messaging App is Truly Secure? 0:00:01.720 --> 0:00:07.440 Also media. 0:00:08.680 --> 0:00:11.800 Welcome back to it could happen here. I am Robert Evans, 0:00:11.880 --> 0:00:16.160 and this is a podcast about things falling apart. Sometimes 0:00:16.160 --> 0:00:19.040 it's about how to make things not fall apart, and 0:00:19.120 --> 0:00:22.360 other times it's more about enduring it. Today is more 0:00:22.400 --> 0:00:24.639 on the endurance side of things. And we're talking about 0:00:24.640 --> 0:00:27.440 a subject that we get a lot of requests about here. 0:00:27.920 --> 0:00:30.240 We've discussed this a year or so ago with one 0:00:30.240 --> 0:00:33.120 of our guests, a great Carl Pasarta. We're talking about 0:00:33.159 --> 0:00:36.839 like security culture, and particularly the aspect of security culture 0:00:36.840 --> 0:00:40.360 that involves digital devices and how to communicate with your friends' 0:00:40.360 --> 0:00:45.920 affinity groups, whomever via your phone essentially or your computer. 0:00:46.400 --> 0:00:49.080 This is a thing where there's a huge amount of 0:00:49.120 --> 0:00:52.599 disinformation as to which apps are safe. What does it 0:00:52.640 --> 0:00:55.960 actually mean to say that an app is encrypted? How 0:00:56.000 --> 0:00:58.840 far does encryption get you? What sort of like cultural 0:00:58.880 --> 0:01:03.920 things come alongside the actual, like physical reality of the 0:01:03.920 --> 0:01:06.160 security of the device in order to kind of make 0:01:06.319 --> 0:01:09.560 a comprehensive security profile. We're gonna be talking about all 0:01:09.560 --> 0:01:11.560 that today and hopefully giving you some good advice on 0:01:11.560 --> 0:01:14.240 what you can trust. Because I am the furthest thing 0:01:14.280 --> 0:01:17.360 in the world from a technical expert. We have two 0:01:17.440 --> 0:01:21.959 actual experts with us today. Carolyn Senders and Cooper Quinton 0:01:22.200 --> 0:01:25.800 have both recently published a paper alongside several other authors 0:01:26.560 --> 0:01:31.360 Leila Wagner, Tim Bernard, Ami Meta, and Justin Hendricks called 0:01:31.400 --> 0:01:34.720 what Is Secure? An Analysis of Popular Messaging Apps, and 0:01:34.760 --> 0:01:38.720 it's it's basically going over what is the actual level 0:01:38.760 --> 0:01:42.120 of security with a number of things like Telegram, you know, 0:01:42.200 --> 0:01:47.880 Telegram's private messaging system, Facebook Messenger, Apple Message, or I Message. 0:01:47.880 --> 0:01:51.200 I guess it's called and obviously signal and kind of 0:01:51.240 --> 0:01:54.080 as a spoiler, signal is your best bet, but that 0:01:54.200 --> 0:01:56.600 also isn't where you should end, right I think we 0:01:56.640 --> 0:01:58.760 want to also talk about kind of like why and 0:01:59.160 --> 0:02:01.680 to what extent that's the case. But anyway, I'm going 0:02:01.720 --> 0:02:04.520 to turn things over to Carolyn and Cooper now because 0:02:04.520 --> 0:02:07.840 I have talked enough about this. Hey guys, welcome to 0:02:07.880 --> 0:02:08.200 the show. 0:02:09.680 --> 0:02:11.760 Hey, Robert, thanks so much for having us on. 0:02:12.280 --> 0:02:15.119 Yeah ah, yeah, thank you so much, A big fan 0:02:15.360 --> 0:02:18.840 of the podcast, so always lovely, really lovely to be here. 0:02:19.320 --> 0:02:20.359 Yeah, thank you so much. 0:02:20.960 --> 0:02:23.840 Yeah, it's really lovely to have you both again. Listeners, 0:02:23.880 --> 0:02:26.600 if you want to take a look at this their paper, 0:02:26.600 --> 0:02:29.000 if you just google what is secure and analysis of 0:02:29.040 --> 0:02:32.000 popular messaging apps. You'll find the Tech Policy Press has 0:02:32.240 --> 0:02:34.760 a summary of it that's pretty quick. The full paper 0:02:34.840 --> 0:02:37.720 is eighty six pages or so. I also recommend reading that, 0:02:37.800 --> 0:02:40.240 but if you wanted to give this the summary a 0:02:40.240 --> 0:02:43.200 skin before you continue, that might help. But I kind 0:02:43.200 --> 0:02:45.760 of wanted to start by asking you, guys, what is 0:02:45.840 --> 0:02:49.600 it that makes Signal a good option for people? Right? 0:02:49.600 --> 0:02:52.040 Because I think most folks you describe it as sort 0:02:52.040 --> 0:02:55.240 of security folklore, right, the stuff that you hear about 0:02:55.280 --> 0:02:58.320 security from your friends, and if you're not a technical person, 0:02:58.360 --> 0:03:00.680 you kind of just like trust what the folks around 0:03:00.720 --> 0:03:02.200 you were saying. And that was sort of how I 0:03:02.240 --> 0:03:05.040 got into Signal. Right, I'm not a technical person, but 0:03:05.080 --> 0:03:07.240 people I knew and trusted who were were like, this 0:03:07.280 --> 0:03:08.080 is your best option? 0:03:08.760 --> 0:03:09.639 Yeah, thank you so much. 0:03:09.760 --> 0:03:11.840 That's such a good question, and I think Cooper and 0:03:11.880 --> 0:03:15.480 I probably have similar but also like very different answers 0:03:15.560 --> 0:03:15.880 to it. 0:03:16.360 --> 0:03:18.320 Cooper, I can go first if you want. 0:03:18.960 --> 0:03:20.560 One of the things I love about Signal is it's 0:03:20.560 --> 0:03:24.200 just really easy to use. It's in and encrypted, it's 0:03:24.200 --> 0:03:26.640 a messaging app. There's not a lot of stuff on it, 0:03:26.680 --> 0:03:29.360 but you can do a lot with it, so you 0:03:29.360 --> 0:03:32.720 can do video calls, you can send actually pretty large 0:03:32.720 --> 0:03:33.799 files like PDFs. 0:03:34.120 --> 0:03:35.680 You can have drag and drop stuff. 0:03:36.080 --> 0:03:39.720 It's like such a low threshold for use for users 0:03:40.040 --> 0:03:42.160 because it is a messaging app, but it does so 0:03:42.200 --> 0:03:45.280 many different kinds of things. But then related to that, 0:03:45.320 --> 0:03:48.920 it's also actually quite minimal. So the paper which everyone 0:03:48.960 --> 0:03:52.120 should read and we'll probably get into this later. Different 0:03:52.160 --> 0:03:58.800 apps like Telegram or Facebook's Messenger app, for example, have have. 0:03:58.840 --> 0:04:00.840 This thing we've been calling feature bloat. 0:04:01.360 --> 0:04:04.160 They are messaging services that actually feel a bit more 0:04:04.240 --> 0:04:06.480 like social networks if you look at the amount of 0:04:06.520 --> 0:04:08.840 stuff that's on there, and by stuff, I don't just 0:04:08.920 --> 0:04:11.880 mean like stickers, I mean if you look at there's 0:04:11.920 --> 0:04:15.080 all these sort of specific and strange settings you can 0:04:15.160 --> 0:04:17.920 use to have all different kinds of messages and all 0:04:17.920 --> 0:04:21.120 different kinds of privacy settings, and all privacy settings are 0:04:21.160 --> 0:04:25.640 really really great. Because Telegram and Facebook Messenger are not 0:04:25.960 --> 0:04:28.840 encrypted by default, actually some of those settings can make 0:04:28.880 --> 0:04:31.279 you feel more secure when you're not so. Kind of 0:04:31.279 --> 0:04:33.159 the beauty of Signal is that out of the box, 0:04:33.680 --> 0:04:37.279 it's incredibly secure. It's an inn encrypted They're not holding 0:04:37.279 --> 0:04:40.360 any data about you. I believe the only only day 0:04:40.400 --> 0:04:42.920 they hold is like when you've like when a phone 0:04:43.000 --> 0:04:46.120 number or a profile has downloaded signal, like when you've 0:04:46.279 --> 0:04:50.600 when you've signed up. But again it's it's incredibly easy 0:04:50.640 --> 0:04:54.159 to use. And another thing is, you know, if this 0:04:54.240 --> 0:04:56.039 was a few years ago, we've been looking at wire 0:04:56.120 --> 0:04:56.679 for example. 0:04:56.680 --> 0:04:58.440 One of the nice things about Signal. 0:04:58.200 --> 0:05:00.800 And this might be controversial to some, is that it 0:05:00.839 --> 0:05:03.640 does follow modern design patterns and standards. So if you're 0:05:03.720 --> 0:05:07.240 using an iOS or Android version, like there are buttons 0:05:07.279 --> 0:05:09.880 in places where you expect them to be. Signal is 0:05:09.880 --> 0:05:15.040 not perfectly designed, but it is quite usable. Yeah, So 0:05:15.080 --> 0:05:17.240 for me, that's kind of what I think makes it 0:05:17.279 --> 0:05:18.400 makes it really wonderful. 0:05:18.680 --> 0:05:21.840 Yeah, it's definitely as much as I love it, and 0:05:21.880 --> 0:05:24.880 it's my like standard messaging app I do every now 0:05:24.920 --> 0:05:26.559 and then run into the thing where like my friends 0:05:26.560 --> 0:05:28.640 will call me through Signal, which is great if you 0:05:28.640 --> 0:05:31.160 need a call to be secure, but it's not nearly 0:05:31.200 --> 0:05:33.280 as good, Like it drops a lot more often than 0:05:33.320 --> 0:05:35.720 a regular phone call, and I'm like, we're just trying 0:05:35.760 --> 0:05:37.680 to meet at the movie theater. It's okay if the 0:05:37.800 --> 0:05:40.800 nsay noes right, Like I've. 0:05:40.640 --> 0:05:42.800 Definitely had that with friends where I'm like, I'm like yeah, 0:05:42.800 --> 0:05:46.640 I'm like, we're just calling to talk about like your dog. 0:05:47.120 --> 0:05:48.320 It's probably fine. 0:05:48.560 --> 0:05:50.839 Yeah, the FBI can have this stuff. 0:05:51.160 --> 0:05:54.800 Yeah, please send, please send, please send dog picks through 0:05:55.279 --> 0:05:56.479 all all messaging apps. 0:05:57.760 --> 0:06:00.839 You know. But on that note, it's uh, right, writing 0:06:01.080 --> 0:06:05.480 usable software that is also secure is really hard, right, 0:06:05.520 --> 0:06:09.560 And like as a like as cryptographer, I'm not a cryptographer, 0:06:09.640 --> 0:06:13.479 but like as somebody cryptographer adjacent, we got that wrong 0:06:13.600 --> 0:06:17.480 for a long time, right, Like before Signal the problem, 0:06:17.680 --> 0:06:20.359 you know, there were the the the sort of most 0:06:20.520 --> 0:06:24.919 used encryption methods were probably uh PGP email, which is 0:06:24.920 --> 0:06:28.159 a method for encrypting email, and off the record chats, 0:06:28.240 --> 0:06:31.760 and both of those none of those ever got to 0:06:31.920 --> 0:06:35.320 the sort of level of user base that Signal and 0:06:35.640 --> 0:06:40.240 and certainly not WhatsApp have, right, And and that's largely 0:06:40.320 --> 0:06:44.880 because they were pretty much unusable, like PGP, almost entirely 0:06:44.960 --> 0:06:49.360 unusable even by cryptography professionals, right, even by computer security 0:06:49.360 --> 0:06:54.080 professionals like ourselves. OTR chat total pain in the butt, right, 0:06:54.200 --> 0:06:58.679 like just just a real nightmare to use. So, like Signal, 0:06:58.880 --> 0:07:01.239 there are still some rough edges, and we talked about 0:07:01.240 --> 0:07:04.280 some of those in our paper. But overall, I think 0:07:04.320 --> 0:07:06.560 that the big the big innovation they've had is just 0:07:07.200 --> 0:07:10.000 remembering that what people want to do on a chat 0:07:10.040 --> 0:07:12.840 app is not encrypt things. What people want to do 0:07:12.880 --> 0:07:14.800 on a chat app is they want to they want 0:07:14.800 --> 0:07:18.040 to chat right. And and the second that that that 0:07:18.120 --> 0:07:20.320 the security sort of gets in the way of that, 0:07:20.880 --> 0:07:23.000 people will stop using it and go find something that's 0:07:23.040 --> 0:07:26.000 more usable. And it seems like that's been Signals sort 0:07:26.040 --> 0:07:30.320 of guiding star and it's and they've you know, doing 0:07:30.360 --> 0:07:33.040 the doing the most secure thing that you can will 0:07:33.120 --> 0:07:38.720 still being fun and usable to actually just chat on right. 0:07:38.800 --> 0:07:40.960 And I think that that has served them quite well. 0:07:41.800 --> 0:07:45.840 Yeah, I think there's it's it's so important. One of 0:07:45.880 --> 0:07:47.960 I think one of the things that that contributes to 0:07:48.040 --> 0:07:52.400 good overall security is setting yourself up for success, which 0:07:52.440 --> 0:07:54.960 means setting yourself up for a system that can function 0:07:55.040 --> 0:07:56.920 well if you're lazy, which is one of the nice 0:07:56.960 --> 0:07:58.800 things that you know, with Signal, you don't have to 0:07:58.840 --> 0:08:01.120 worry about like opting in and out and like selecting 0:08:01.120 --> 0:08:03.840 a bunch of stuff. It's pretty safe, especially for a 0:08:03.840 --> 0:08:07.040 normal person's uses right out of the box, which is 0:08:07.160 --> 0:08:10.040 huge and kind of in the same line, as that 0:08:10.160 --> 0:08:13.200 is the fact that because Signal doesn't store metadata, you're 0:08:13.200 --> 0:08:16.760 not relying upon them being like committed you know, anti 0:08:16.760 --> 0:08:20.440 state actors or whatever like, because they don't have access 0:08:20.480 --> 0:08:22.840 to the thing that for example, Facebook will hand over 0:08:22.920 --> 0:08:25.560 to the cops if the cops just like breathe in 0:08:25.600 --> 0:08:26.160 their direction. 0:08:28.040 --> 0:08:30.680 Yeah, that's that's exactly right, And that's that is that 0:08:30.760 --> 0:08:32.760 is the other really cool thing about Signal. You know, we, 0:08:33.400 --> 0:08:36.560 as Carolyn said, the only data that Signal gives over 0:08:36.640 --> 0:08:40.480 in response to uh A subpoena is the time that 0:08:41.040 --> 0:08:43.319 the phone number signed up for Signal account and the 0:08:43.400 --> 0:08:45.760 last time it connected to the Signal server. And the 0:08:45.800 --> 0:08:49.600 reason we know that is because Signal publishes transparency reports 0:08:50.080 --> 0:08:53.199 with the full text and full response of any subpoena 0:08:53.240 --> 0:08:56.480 that they get, so like we can actually just see 0:08:56.760 --> 0:08:59.280 in the responses that all they've given over is these 0:08:59.320 --> 0:09:01.920 two pieces of information, because that's all they have, and 0:09:01.960 --> 0:09:05.319 they've done some pretty clever things to make that be the. 0:09:05.280 --> 0:09:08.840 Case, right, And that's actually so different than how other 0:09:08.960 --> 0:09:13.160 companies are I think reporting on either subpoenas or any 0:09:13.200 --> 0:09:13.640 kind of. 0:09:15.280 --> 0:09:17.040 Weight that law enforcement puts on them. 0:09:17.400 --> 0:09:20.160 So for our report, I don't remember how much it's 0:09:20.200 --> 0:09:23.120 it's mentioned in the report actually, but we did go 0:09:23.240 --> 0:09:28.440 through and look at Apple Meta and I think Google 0:09:28.520 --> 0:09:31.040 like in their own transparency reports to try to get 0:09:31.040 --> 0:09:34.000 a sense of how how that would stack up in 0:09:34.040 --> 0:09:38.120 comparison to Signals. I think in some cases it's saying 0:09:38.280 --> 0:09:42.640 like they received some kind of like notification, but like no, 0:09:43.880 --> 0:09:47.840 nothing really clear or specific on like what they received 0:09:47.880 --> 0:09:50.640 from law enforcement or government, but rather just that they 0:09:50.720 --> 0:09:52.640 received one. And so that's also the really great thing 0:09:52.679 --> 0:09:55.600 about Signal is you are getting all of this information 0:09:56.000 --> 0:10:00.400 that you're not getting from other companies or PLATF forms. 0:10:01.040 --> 0:10:04.400 Yeah, you know, I wanted to kind of in the 0:10:04.480 --> 0:10:07.280 same subject and going back to we kind of opened 0:10:07.280 --> 0:10:11.320 this introducing the concept that y'all introduced me to. I 0:10:11.320 --> 0:10:13.480 guess I was aware of this, but not the terminology 0:10:13.520 --> 0:10:16.080 security folklore, and I wanted to chat a little bit 0:10:16.080 --> 0:10:18.040 about kind of the most recent example of this something 0:10:18.320 --> 0:10:20.560 a lot of folks have probably been wondering about since 0:10:20.559 --> 0:10:23.200 we started talking about Signal, which is that roughly a 0:10:23.240 --> 0:10:27.320 week before y'all and I sat down to talk about this, 0:10:28.320 --> 0:10:32.280 a kind of viral info meme started coming through that 0:10:32.400 --> 0:10:36.200 was like Signal has a zero day exploit, which is 0:10:36.559 --> 0:10:40.320 basically a hole that a hacker found in an Apple program. 0:10:40.520 --> 0:10:43.880 That is that can't expose you. You have to turn off 0:10:45.160 --> 0:10:47.720 link previews, right, which is that when you when someone 0:10:47.760 --> 0:10:50.040 sends you like a link to an article in Signal, 0:10:50.240 --> 0:10:53.800 you get a little preview not dissimilar to how it is. 0:10:54.360 --> 0:10:56.720 And I think to be fair, just based on my 0:10:56.880 --> 0:10:59.559 very limited knowledge, that is, when I think about, like, 0:10:59.600 --> 0:11:01.840 what are potential holes in Signal, I don't think it's 0:11:01.880 --> 0:11:05.680 unreasonable to be concerned about that specific feature. But that 0:11:05.920 --> 0:11:09.040 warning was not what it kind of seemed to be basically, 0:11:09.120 --> 0:11:10.679 or not as accurate as I think a lot of 0:11:10.720 --> 0:11:11.760 people took it as being. 0:11:11.800 --> 0:11:12.080 I don't know. 0:11:12.160 --> 0:11:13.840 I'll let I'll tell I'll turn it over to you, guys. 0:11:13.840 --> 0:11:15.280 I think that's the next thing I want to talk about. 0:11:15.600 --> 0:11:18.480 I'll turn it over to Cooper, who had you had 0:11:18.679 --> 0:11:21.600 a Uh, you have a lot of feels about that. 0:11:21.880 --> 0:11:25.640 I have so many feelings about this. I was working 0:11:25.679 --> 0:11:26.680 on this all weekend. 0:11:26.800 --> 0:11:30.520 So this yeah, so this copy pasta I'm calling this 0:11:30.600 --> 0:11:33.800 like this signal copy pasta yeah, which is a term 0:11:33.880 --> 0:11:37.960 from you know, four Chan and other horrible internet places, 0:11:37.960 --> 0:11:39.640 but some. 0:11:40.120 --> 0:11:42.440 Media audience is probably Internet enough. 0:11:43.240 --> 0:11:45.600 Yeah, I'm gonna guess a good half of the people 0:11:45.640 --> 0:11:47.880 listening at least got that message. 0:11:47.920 --> 0:11:52.080 Yeah yeah. And it's it's like, first of all, this 0:11:52.240 --> 0:11:55.760 is not if you if you had a zero day 0:11:55.760 --> 0:11:57.560 in Signal, which is it, which is an exploit for 0:11:57.600 --> 0:12:00.360 Signal that has been unpatched but has not been patched 0:12:00.360 --> 0:12:02.560 by the vendor so you can actively exploit it. There 0:12:02.600 --> 0:12:06.360 are no people in the world who would choose to 0:12:08.520 --> 0:12:12.760 quietly leak this over you know, over vague signal texts. 0:12:13.000 --> 0:12:15.960 There are two types of people. One uh, you know, 0:12:16.200 --> 0:12:19.559 people like us that would bring this to Signal immediately 0:12:19.600 --> 0:12:22.320 and get them to patch it to protect the you know, 0:12:22.400 --> 0:12:25.079 millions of high risk users that you signal, or to 0:12:25.360 --> 0:12:27.520 the type of people that would go sell this exploit 0:12:27.960 --> 0:12:30.440 to some horrible company that would use it, you know, 0:12:30.520 --> 0:12:32.920 sell it to to Saudi Arabia or something and use 0:12:32.960 --> 0:12:35.800 it to kill activists. Right, Like there is and there's 0:12:35.840 --> 0:12:39.160 no in between. There's nobody that is going to quietly 0:12:39.240 --> 0:12:43.840 leak this for you know, just for fun with vague details. Right. So, 0:12:43.840 --> 0:12:46.640 so this this message set up red flags immediately, and 0:12:47.480 --> 0:12:51.280 like it's because I really do not like lying previews, 0:12:51.320 --> 0:12:54.000 And in our paper we discussed some of the issues 0:12:54.040 --> 0:12:56.800 that we have with link previews. You know, we think 0:12:56.840 --> 0:13:00.360 that they can they can leak some information about your 0:13:00.840 --> 0:13:04.319 chats to the owner of the website. Right. We think 0:13:04.360 --> 0:13:06.680 it's a kind of a large attack service. It's not 0:13:06.800 --> 0:13:07.719 super necessary. 0:13:08.120 --> 0:13:11.640 Would you mind explaining to actually the audience to like 0:13:12.240 --> 0:13:15.840 a little bit about what what we found when looking 0:13:15.880 --> 0:13:16.880 at link previews. 0:13:17.520 --> 0:13:20.880 Yeah. So, the way that link previews work is when 0:13:20.920 --> 0:13:23.719 you the way that they work on Signal and on 0:13:23.760 --> 0:13:27.480 WhatsApp is that when you send a link to somebody, 0:13:28.120 --> 0:13:32.400 the Signal app or WhatsApp goes and like fetches the 0:13:32.440 --> 0:13:34.520 web page that that you know for that link, Right, 0:13:34.559 --> 0:13:37.000 It goes and downloads, you know, downloads the content of 0:13:37.040 --> 0:13:41.079 that link and gets a There are some there's some 0:13:41.120 --> 0:13:45.080 special HTML tags that describe, you know, sort of what 0:13:45.120 --> 0:13:47.200 the page is about, what the title of the page is, 0:13:47.240 --> 0:13:49.400 and like an image for the page. And it gets 0:13:49.480 --> 0:13:51.280 those tags and it puts them all together in this 0:13:51.280 --> 0:13:53.839 little package and then sends that all as part of 0:13:53.920 --> 0:13:57.120 the signal message. So when you put a link in Signal, 0:13:57.160 --> 0:13:59.120 your phone actually goes out and gets that web page, 0:13:59.120 --> 0:14:04.360 and it gets that web page with a what's called 0:14:04.360 --> 0:14:07.040 the user agent, which is like a piece of text 0:14:07.080 --> 0:14:11.079 that's attached to the request that uniquely that that identifies 0:14:11.120 --> 0:14:15.280 it as being a request from Signal and from like 0:14:15.320 --> 0:14:18.680 from signal and from your IP address. Right, So when 0:14:18.679 --> 0:14:20.880 you put a link in, the owner of that website, 0:14:20.880 --> 0:14:23.200 whoever has the logs for that website can know that 0:14:24.280 --> 0:14:27.240 somebody at your IP address is using signal and sending 0:14:27.320 --> 0:14:31.440 this link over signal. What we're what our concern is 0:14:31.440 --> 0:14:32.360 is that if that. 0:14:32.280 --> 0:14:37.960 Link is unique, then anybody else who visits that link 0:14:38.400 --> 0:14:42.560 can be inferred to be somebody that you are talking 0:14:42.600 --> 0:14:43.280 with over. 0:14:43.200 --> 0:14:47.720 Signal, right, And so like this can be this can 0:14:47.760 --> 0:14:51.920 be a good an interesting a source of intelligence for 0:14:53.040 --> 0:14:56.680 website owners, especially for big websites that can easily generate 0:14:56.760 --> 0:14:59.800 unique links with like tracking parameters at the end of 0:14:59.800 --> 0:15:04.320 the right, Like when you share a Instagram post and 0:15:04.560 --> 0:15:06.560 like at the end it's like question mark I G 0:15:06.920 --> 0:15:09.320 S H I D equals you know, a long string 0:15:09.360 --> 0:15:12.200 of numbers and letters, right, or a Twitter post where 0:15:12.320 --> 0:15:15.160 you know T equals a long string of letters and numbers. Right. 0:15:15.240 --> 0:15:17.880 That makes a unique link, and then anybody who visits 0:15:17.920 --> 0:15:20.880 that same link can be determined to be somebody that 0:15:20.920 --> 0:15:27.400 you're speaking with over Signal, so and also WhatsApp and so. 0:15:27.400 --> 0:15:33.080 So for that reason, we we we think that Signal 0:15:33.160 --> 0:15:36.640 and WhatsApp should turn link previews off by by default 0:15:37.520 --> 0:15:40.800 because we think that that's an unncessary information. Link Signal 0:15:40.840 --> 0:15:45.560 and WhatsApps pushed back on that is that link previews 0:15:45.640 --> 0:15:51.040 are a core feature that people demand and if they 0:15:51.200 --> 0:15:53.440 if they were to turn off link previews by default, 0:15:53.960 --> 0:15:56.800 they're worried that people would leave the platform for less 0:15:56.800 --> 0:15:59.360 secure platforms like Telegram. 0:16:00.480 --> 0:16:00.640 Yeah. 0:16:00.640 --> 0:16:02.600 I mean, I don't want to tell them their business, 0:16:02.600 --> 0:16:05.920 because I'm sure they have data on this, but I've 0:16:06.000 --> 0:16:10.920 never thought about link previews as being a thing that 0:16:11.000 --> 0:16:11.800 I needed. 0:16:12.920 --> 0:16:15.480 It's like, yeah, I think it's I think it's one 0:16:15.520 --> 0:16:18.560 of those things. And you know, we haven't necessarily done 0:16:18.840 --> 0:16:24.120 like extensive general design research in this right, Like we 0:16:24.200 --> 0:16:27.600 haven't surveyed like three thousand people in the US. We 0:16:27.640 --> 0:16:31.080 haven't had like a Pew Research survey across countries and 0:16:31.200 --> 0:16:32.880 be like, what are your thoughts on link previews? 0:16:33.760 --> 0:16:34.920 But I would. 0:16:34.600 --> 0:16:38.400 Probably argue because it is it is included in so 0:16:38.560 --> 0:16:40.200 much of modern messaging. 0:16:39.800 --> 0:16:43.000 Apps that we now assume it's like a core feature. 0:16:43.840 --> 0:16:45.880 One thing I will give Signal that I think is 0:16:45.920 --> 0:16:48.760 amazing that other apps don't do, and this is true 0:16:48.800 --> 0:16:52.800 of WhatsApp is pretty much every feature except for encryption, 0:16:52.920 --> 0:16:55.920 you can there's something you can toggle or turn off. Right, 0:16:56.000 --> 0:17:00.440 So like link preview already was available for people to 0:17:00.440 --> 0:17:05.080 turn off on Signal, WhatsApp does not allow that, and 0:17:05.840 --> 0:17:09.920 it seems like they're making no moves to allow that 0:17:09.960 --> 0:17:12.680 future to be optional to turn on or off. 0:17:13.000 --> 0:17:13.880 But that is I will say. 0:17:13.880 --> 0:17:15.639 One of the things that's really lovely about Signal that 0:17:16.080 --> 0:17:20.680 is so different from modern design and modern like big 0:17:20.720 --> 0:17:23.480 tech platforms and just platforms in general, is that those 0:17:23.640 --> 0:17:26.720 a lot of features are optional, whereas you know, WhatsApp 0:17:26.720 --> 0:17:29.399 in metas sort of stance on design is that a 0:17:29.400 --> 0:17:31.920 lot of things are not optional, that those are things 0:17:32.000 --> 0:17:35.439 users would want. Why would we make foundational elements like 0:17:35.480 --> 0:17:40.359 link previews optional? And you're just like starting like gesturing wildly, 0:17:40.480 --> 0:17:42.560 but like you know, it's like, well, you don't know 0:17:42.560 --> 0:17:44.720 what people want, and I mean, what's the harm in 0:17:44.760 --> 0:17:46.320 turning off some of some of these things? 0:17:46.400 --> 0:17:46.560 Right? 0:17:47.200 --> 0:17:50.320 You know, like maybe maybe people don't want to receive gifts. 0:17:50.400 --> 0:17:52.320 I don't know, maybe they don't want to receive stickers. 0:17:52.359 --> 0:17:55.040 Why don't you like let them have that option. What's 0:17:55.040 --> 0:17:56.000 the harm that could happen? 0:17:56.200 --> 0:17:56.440 Yeah? 0:17:56.720 --> 0:17:58.520 Yeah, yeah, I couldn't agree more. 0:17:58.560 --> 0:18:01.119 Yeah. Two things I want to say that one is 0:18:01.240 --> 0:18:03.760 one is that and first we should acknowledge that this 0:18:03.920 --> 0:18:05.840 it turns out that there was no zero day, there 0:18:05.880 --> 0:18:10.040 was no vulnerability. Yeah, this was absolutely just something that 0:18:10.040 --> 0:18:13.960 that spread virally out of nowhere. I'd be really interested 0:18:13.960 --> 0:18:16.600 to find out what the origin of this copy of 0:18:16.640 --> 0:18:19.119 past I was, but I haven't. I haven't been able to. 0:18:19.200 --> 0:18:20.639 But it's I'm. 0:18:20.520 --> 0:18:21.840 Curious about that as as well. 0:18:21.880 --> 0:18:23.560 Because I was in another group threw that was like, 0:18:23.600 --> 0:18:25.960 we really need outside auditors to look at these. 0:18:25.800 --> 0:18:27.840 And I was like, we have a whole report that 0:18:27.960 --> 0:18:29.919 we wrote that didn't look at this. 0:18:30.920 --> 0:18:34.840 Speaking of outside auditors, I gotta pause you guys just 0:18:34.840 --> 0:18:37.760 a second, because it is time for an ad break, 0:18:38.400 --> 0:18:42.600 So please spend your money and then come back to 0:18:42.680 --> 0:18:59.080 learn more. Ah and we're back. Okay, sorry about that, Cooper, Carolyn, 0:18:59.640 --> 0:19:02.000 you make continue as you were. 0:19:02.400 --> 0:19:04.480 The other thing I was, I was going to say 0:19:05.080 --> 0:19:09.520 that the idea that anybody would leave WhatsApp because they 0:19:09.560 --> 0:19:13.800 stopped having link previews is completely preposterous to me. Like 0:19:14.080 --> 0:19:21.600 Clownish has over two billion users. They are the you know, 0:19:21.720 --> 0:19:26.240 in a position to set the standard for what people 0:19:26.280 --> 0:19:31.440 expect from a messaging app, and so like they could 0:19:31.520 --> 0:19:35.840 do things like turn on disappearing messages by default and 0:19:35.920 --> 0:19:38.320 change that culture. They could do things like turn off 0:19:38.320 --> 0:19:41.440 link previews by default and change that culture. Like, they 0:19:41.440 --> 0:19:44.159 could do these things, and you know they would you know, 0:19:45.280 --> 0:19:50.600 they would not lose enough users to even notice or 0:19:50.640 --> 0:19:51.159 care about. 0:19:51.320 --> 0:19:51.439 Right. 0:19:51.520 --> 0:19:53.760 Yeah, they are the only people in the position in 0:19:53.800 --> 0:19:57.080 the world, in the position to decide what the culture 0:19:57.119 --> 0:19:59.159 should be, and this is what they've decided the culture 0:19:59.200 --> 0:19:59.600 should be. 0:20:00.240 --> 0:20:00.680 Totally. 0:20:01.040 --> 0:20:03.040 I hate to break it to you, but if WhatsApp 0:20:03.320 --> 0:20:05.879 just got rid of link previews, I'm just throwing my 0:20:05.960 --> 0:20:09.040 whole phone into the garbage garbage can, getting rid of it. 0:20:09.040 --> 0:20:11.159 Just tossing it back to a landline. 0:20:11.560 --> 0:20:12.080 Yeah, I'm just. 0:20:12.000 --> 0:20:14.240 Gonna eat it into a river. I feel like I 0:20:14.280 --> 0:20:17.360 don't need this anymore. Actually, I'm going back to carryer pigeons. 0:20:17.400 --> 0:20:18.919 That's how far back I'm going to go. 0:20:19.280 --> 0:20:21.280 I mean that that does kind of lead into the 0:20:21.280 --> 0:20:23.480 next thing I wanted to talk about, which is sort 0:20:23.520 --> 0:20:28.760 of the other wing from security folklore, which is security nihilism. 0:20:29.400 --> 0:20:32.160 And yeah, this is kind of you introduce this when 0:20:32.200 --> 0:20:35.400 talking about sort of if you do try to engage 0:20:35.480 --> 0:20:37.760 somewhat with the technology, or if you wind up just 0:20:37.800 --> 0:20:39.760 kind of in the position I think most lay people are, 0:20:39.840 --> 0:20:42.240 where you know, maybe you have some friends who know more, 0:20:42.440 --> 0:20:44.200 or maybe you have some friends who think they know more, 0:20:44.240 --> 0:20:46.760 and you get all these conflicting things about like this 0:20:46.920 --> 0:20:49.159 is safe, No, it's not. You can't trust signal. The 0:20:49.160 --> 0:20:51.280 FEDS could be running signal all this kind of stuff, 0:20:51.880 --> 0:20:54.560 And to be fair, the FEDS have run security based 0:20:54.600 --> 0:20:57.159 services before. It's not like I don't believe that's happening 0:20:57.160 --> 0:20:59.920 with signal, but it's not like I understand where parent 0:21:00.480 --> 0:21:04.239 like that can can enter into people's calculus, especially if 0:21:04.280 --> 0:21:08.560 you're not technically knowledgeable, and that can lead to this 0:21:08.640 --> 0:21:11.959 sort of state of security nihilism where you're just like, 0:21:12.400 --> 0:21:14.720 you can't communicate it all online. There's no way to 0:21:14.760 --> 0:21:18.199 do it securely, and obviously there's no perfect right you 0:21:18.240 --> 0:21:20.040 never have it, but you don't have one hundred percent 0:21:20.119 --> 0:21:24.760 with like talking in person to somebody. Right there are 0:21:24.760 --> 0:21:28.280 individuals in prison right now who you know somebody they 0:21:28.359 --> 0:21:31.240 loved and trusted rat it on them. There's no one 0:21:31.280 --> 0:21:35.399 hundred percents in this world. But that doesn't mean nihilism 0:21:35.480 --> 0:21:38.280 is the right response to like trying to figure out 0:21:38.400 --> 0:21:41.920 how to set up your communications standards with people right. 0:21:42.320 --> 0:21:44.680 Totally, I mean, I think the approach we take in 0:21:44.920 --> 0:21:48.399 because throughout this report we were also teaching workshops to 0:21:49.040 --> 0:21:53.120 reproductive justice activists across the US and states where abortion 0:21:53.240 --> 0:21:56.680 is banned. I'm from Louisiana, I live half the year there, 0:21:57.040 --> 0:22:01.600 the abortion is banned there, and we were also working 0:22:01.600 --> 0:22:04.000 with journalists in India. So a big big thing for 0:22:04.080 --> 0:22:07.520 us was also teaching threat modeling and different kinds of 0:22:07.680 --> 0:22:11.359 what Matt Mitchell, a security trainer and expert, calls digital hygiene, 0:22:11.840 --> 0:22:14.840 and so a lot of this was recognizing that there 0:22:14.960 --> 0:22:18.000 was certain practices we were picking up on, particularly with 0:22:18.119 --> 0:22:19.639 folks we were working with. So like a lot of 0:22:19.720 --> 0:22:23.399 reproductive justice activists we were working with are new to security, 0:22:23.440 --> 0:22:25.720 they're new to technology, they don't have a background in tech, 0:22:25.840 --> 0:22:29.880 and generally, you know, the American South, the American Deep 0:22:29.920 --> 0:22:33.520 South is super overlooked in terms of tech policy, in 0:22:33.640 --> 0:22:37.200 terms of just I think a general focus when people 0:22:37.280 --> 0:22:40.280 are talking about tech or tech literacy or tech activism, 0:22:40.960 --> 0:22:44.160 and that is like leaving really massive gaps and knowledge 0:22:44.240 --> 0:22:47.760 for people. And so you know, when we were working 0:22:47.800 --> 0:22:50.720 on this security folkal or and security nihilism, we're both 0:22:50.840 --> 0:22:51.840 actually very. 0:22:51.920 --> 0:22:53.639 Almost like I won't say, like a pendulum, but they 0:22:53.640 --> 0:22:56.199 were very connected. And so some of that was. 0:22:56.200 --> 0:22:58.520 People hearing things like oh, I should put my phone 0:22:58.960 --> 0:23:02.120 in a microwave when I'm having a very sensitive conversation, right, 0:23:02.160 --> 0:23:04.480 And so that's where some of that security folklore is 0:23:04.520 --> 0:23:08.399 coming in. It is something that is technically safe, but 0:23:08.480 --> 0:23:10.920 it's like not the thing you necessarily, like totally need 0:23:10.960 --> 0:23:13.760 to do in that moment. And with security nihilism, what 0:23:13.840 --> 0:23:15.359 it kind of came down to, and this is stuff 0:23:15.359 --> 0:23:20.000 we've seen with other groups and other circumstances. A great 0:23:20.040 --> 0:23:23.520 example are are you know Palestinian activists and journalists. Let's say, 0:23:23.560 --> 0:23:26.680 who are you know facing the threat of all different 0:23:26.760 --> 0:23:30.080 kinds of governmental censorship and surveillance of sort of saying like, 0:23:30.080 --> 0:23:32.560 when there's this large threat sort of hanging on us, 0:23:33.040 --> 0:23:36.439 and there's also physical surveillance. And this is true for 0:23:36.480 --> 0:23:39.120 a lot of journalists in other countries like India as well. 0:23:39.240 --> 0:23:44.280 For example, you know, like should everything go through signal 0:23:44.359 --> 0:23:45.280 or does it really matter? 0:23:45.400 --> 0:23:46.400 Like does it really matter? 0:23:46.480 --> 0:23:49.080 And this is also something again we saw with some 0:23:49.080 --> 0:23:51.479 some reproductive justice activists as well, where it's like if 0:23:51.520 --> 0:23:53.480 everything is being monitored, what's safe? 0:23:53.680 --> 0:23:57.080 Like can I send stuff? Like can I even use Google? 0:23:57.480 --> 0:24:01.480 And part of this was, you know, by teaching privacy 0:24:01.480 --> 0:24:04.680 and security workshops, by teaching things like threat modeling, which 0:24:04.720 --> 0:24:06.720 is a framework for just assessing what. 0:24:06.800 --> 0:24:08.200 Are what are threats? 0:24:08.560 --> 0:24:11.040 Like are what are all the potential threats you could 0:24:11.040 --> 0:24:13.359 face and kind of mapping them from like the most 0:24:13.400 --> 0:24:16.159 minor to like the most major, and what you can 0:24:16.200 --> 0:24:18.320 do about that. That's a way to try to combat 0:24:18.440 --> 0:24:20.840 security nihilism. But I think an approach Cooper and I 0:24:20.880 --> 0:24:23.119 are also really fond of is thinking of this like 0:24:23.160 --> 0:24:26.040 safer sex. There's all different kinds of things you can 0:24:26.080 --> 0:24:29.679 do that our mitigations are actually incredibly helpful, and we 0:24:29.760 --> 0:24:32.240 can't look at it as a binary of safe or 0:24:32.400 --> 0:24:35.600 not safe. It's actually like much more of a gradient. 0:24:37.240 --> 0:24:40.080 But you know, the focal are and the nihilism, I 0:24:40.080 --> 0:24:42.520 think come from a very similar place, which is we're 0:24:42.560 --> 0:24:45.960 asking people like society is kind of asking or demanding 0:24:46.000 --> 0:24:49.880 that people be experts and something that's really hard. I 0:24:49.920 --> 0:24:52.480 am like a fairly technical person, and even there are 0:24:52.560 --> 0:24:54.639 some things that I find hard to serve wrap my 0:24:54.680 --> 0:24:57.320 head around. And I've been working in privacy and security 0:24:57.359 --> 0:25:00.719 for like quite a while, and I think think, you know, 0:25:01.920 --> 0:25:03.879 it's also really hard when you think about these apps 0:25:03.880 --> 0:25:06.280 as like a brand new person. So, like, one of 0:25:06.280 --> 0:25:07.840 the things that popped up a lot in our research 0:25:07.960 --> 0:25:10.080 is like why should we trust signal? And that's actually 0:25:10.160 --> 0:25:13.000 a great question, Like what about Signal in its interface 0:25:13.040 --> 0:25:14.120 and its design. 0:25:14.760 --> 0:25:17.320 Would cause you to trust it? Like some people were 0:25:17.359 --> 0:25:19.359 like it's a nonprofit. That's great, but I don't know 0:25:19.359 --> 0:25:21.640 what that means. I'm like, that's actually a fantastic question, 0:25:21.720 --> 0:25:23.040 Like what does that mean? Right? 0:25:23.080 --> 0:25:26.040 Like why should you trust this? You've heard through the 0:25:26.040 --> 0:25:28.479 grapevine that you should. And I think these are kind 0:25:28.520 --> 0:25:30.240 of all the things that people are dealing with because 0:25:30.240 --> 0:25:32.000 if you sort of take a step back and just 0:25:32.040 --> 0:25:34.920 look at software or any different kinds of software generally, 0:25:35.359 --> 0:25:37.600 why should you trust that it's safe and secure when 0:25:37.640 --> 0:25:40.480 there have been so many different kinds of leaks or 0:25:40.520 --> 0:25:46.040 breaches or things breaking, right, Yeah, Like, so these are 0:25:46.359 --> 0:25:48.600 I think really really closely tied. But I think a 0:25:48.640 --> 0:25:52.080 big thing for us is trying to combat that security nihilism. 0:25:52.200 --> 0:25:54.719 Like when whenever we can like, there is things you 0:25:54.760 --> 0:25:57.480 can do. I don't want to say like no matter 0:25:57.480 --> 0:25:59.560 how great the threat, but I believe like, no matter 0:25:59.600 --> 0:26:01.760 how great threat, there is stuff, there is stuff you 0:26:01.800 --> 0:26:02.120 can do. 0:26:02.760 --> 0:26:04.679 No matter how great the threat is, there's stuff that 0:26:04.720 --> 0:26:07.920 you can do to make it more difficult and more 0:26:07.920 --> 0:26:10.679 expensive for that person to attack you. Right, Like we 0:26:10.760 --> 0:26:13.480 all lock the doors to our house, or you know, 0:26:13.560 --> 0:26:16.639 for the most part, or you know, we all we 0:26:16.720 --> 0:26:20.320 all do things to to protect ourselves like that that 0:26:20.400 --> 0:26:22.680 aren't fool proof, right. Somebody can always break a window 0:26:22.720 --> 0:26:24.680 to get into your house, right. Somebody can find other 0:26:24.680 --> 0:26:27.040 ways to get into your house. But locking the door 0:26:27.240 --> 0:26:30.000 makes it so that somebody has to do the noisy 0:26:30.040 --> 0:26:33.600 thing of breaking a window. Right. It makes it so that, 0:26:33.640 --> 0:26:35.840 you know, somebody has to spend more time and effort 0:26:35.920 --> 0:26:40.000 and more risk of getting caught in getting into your house. Right. 0:26:40.040 --> 0:26:42.760 And that's and that's like we layer. When you layer 0:26:42.880 --> 0:26:47.160 these protections, right, the idea, you know is that you're 0:26:47.280 --> 0:26:51.200 you're you're making it harder, You're making there be more 0:26:51.240 --> 0:26:54.200 friction right to piercing your security. 0:26:54.760 --> 0:26:57.280 Yeah, I think that's that's a really good point, and 0:26:57.320 --> 0:26:59.720 that the concept of friction, you know, this is something 0:26:59.760 --> 0:27:02.639 I've talked about. Not that these are exactly the same things, 0:27:02.640 --> 0:27:05.360 but in the although there's not not wildly different when 0:27:05.359 --> 0:27:09.360 it comes to like how insurgents win insurgencies, right, it's 0:27:09.800 --> 0:27:13.439 not by carrying out these sort of like great battlefield 0:27:13.560 --> 0:27:17.760 victories that sweep the enemy from the field. It's by friction, right, 0:27:18.040 --> 0:27:22.199 which wears down both the culture and the kind of 0:27:22.280 --> 0:27:27.080 readiness of the opponent until they simply bounce, which is 0:27:27.160 --> 0:27:29.679 a pretty durable and effective strategy. You can keep it up. 0:27:29.720 --> 0:27:35.679 It's this matter of like there's no like sweeping sudden 0:27:35.800 --> 0:27:39.719 like ninety minute three act win here. It's more a 0:27:39.720 --> 0:27:43.239 matter of the more difficult, the more expensive you make it, 0:27:43.480 --> 0:27:46.240 the more you hold on to and the more all 0:27:46.280 --> 0:27:48.160 of us hold on to. Right. That's the other benefit 0:27:48.240 --> 0:27:50.200 is like, even if you're not even if you are 0:27:50.400 --> 0:27:53.000 the most law abiding person in the world like myself, 0:27:54.600 --> 0:27:58.160 having these security measures in place means that you're kind 0:27:58.160 --> 0:28:02.480 of contributing to the overall immune system of a of 0:28:02.520 --> 0:28:05.399 a kind of community of people who don't want the 0:28:05.520 --> 0:28:06.720 NSA listening to their ship. 0:28:08.600 --> 0:28:12.680 Yeah, exactly, exactly. And the friction thing is is also 0:28:12.760 --> 0:28:16.800 exactly what Signal does, right, Like by the threat model 0:28:16.840 --> 0:28:21.080 for Signal is stopping the NSA or other global adversaries 0:28:21.119 --> 0:28:25.560 from listening to all communications as they travel over the internet, right, 0:28:25.680 --> 0:28:28.360 And that's when you can when you can do that, 0:28:28.720 --> 0:28:30.840 like when you can when you can listen to everybody's 0:28:30.880 --> 0:28:33.800 conversations as they travel over the Internet, it's really cheap 0:28:33.840 --> 0:28:37.080 to spy on anybody. Right when you're encrypting that communication, 0:28:38.160 --> 0:28:41.239 then the NSA or whatever other global adversary has to 0:28:41.280 --> 0:28:45.120 go actually hack your phone, right, they have to. They 0:28:45.120 --> 0:28:48.920 have to target you specifically, they have to burn resources 0:28:49.000 --> 0:28:53.920 and you know, burn weapons, right, zero days to get 0:28:53.960 --> 0:28:56.080 access to your phone. And that's a lot more costly, 0:28:56.560 --> 0:28:59.240 it's a lot more noisy, it's a much higher risk 0:28:59.280 --> 0:29:02.280 of them getting caught. So it's introduced to huge friction, 0:29:02.920 --> 0:29:08.560 uh in that in that area. 0:29:08.720 --> 0:29:10.200 Go ahead, okay, go ahead, go. 0:29:10.160 --> 0:29:13.040 Ahead, I say, and I think you're asymmetic. The sort 0:29:13.080 --> 0:29:16.000 of comparison to asymmetric warfare is exactly spot on, because 0:29:16.040 --> 0:29:17.840 none of us are ever going to have the money 0:29:17.840 --> 0:29:20.160 that that the NSA or Masade has. None of us 0:29:20.160 --> 0:29:23.760 are ever going to have the the total technical acumen 0:29:23.840 --> 0:29:26.560 that the NSA or MASAD has, right, but like those 0:29:27.200 --> 0:29:28.920 that you know, so we have to kind of fight 0:29:29.000 --> 0:29:32.080 a you know, in terms of caryption, in terms of encryption, 0:29:32.160 --> 0:29:35.320 a guerrilla war, right, and we have to make things 0:29:35.840 --> 0:29:38.840 so expensive and so annoying for them that it's not. 0:29:38.760 --> 0:29:41.800 Worth it totally. And just to sort of building that. 0:29:41.840 --> 0:29:44.000 One of the things I love about Signal is while 0:29:44.040 --> 0:29:48.680 they're creating friction for our adversaries, it's actually so frictionless 0:29:48.720 --> 0:29:51.440 to use as a user. And I think that's one 0:29:51.480 --> 0:29:54.800 of the things I find just continually impressive about that. 0:29:55.120 --> 0:29:58.160 I don't want this to turn into like the like. 0:29:58.160 --> 0:30:01.200 We're all himbos for signal. Looks we probably are. 0:30:01.720 --> 0:30:03.760 But because like that's one of the things as a 0:30:03.800 --> 0:30:06.200 researcher like Kuber and IOMs have to be like, we're 0:30:06.280 --> 0:30:08.640 not paid by a signal at all, Like, but this 0:30:08.840 --> 0:30:10.640 is in fact, like one of the best things you 0:30:10.680 --> 0:30:12.480 can use. But again, one of the things I think 0:30:12.520 --> 0:30:16.000 is amazing is that it is so easy to use 0:30:17.080 --> 0:30:21.440 and it really is designed for and I'm using the 0:30:21.520 --> 0:30:25.960 term usability as as a design term, meaning that it 0:30:26.000 --> 0:30:29.920 is they're thinking about a common user, including those with 0:30:29.960 --> 0:30:33.480 like lower digital literacy or those that are have never 0:30:33.680 --> 0:30:37.160 used any kind of any kind of security tool, and 0:30:37.240 --> 0:30:40.800 so they're hitting a specific threshold of usability for things 0:30:40.840 --> 0:30:44.120 to be understandable. And again, that's incredibly hard to do well, 0:30:44.280 --> 0:30:46.280 and they are they are doing it quite well. Like 0:30:46.320 --> 0:30:48.760 it's very I would argue, it's very easy and sort 0:30:48.760 --> 0:30:52.080 of seamless for people to make a jump from WhatsApp 0:30:52.240 --> 0:30:55.240 or if you're on like Google or Android using like 0:30:55.280 --> 0:30:59.400 Google Messages, sorry Google, if you're on Android or an iPhone, 0:30:59.600 --> 0:31:03.200 from Like Messages to Google Messages to signal like it 0:31:03.240 --> 0:31:05.400 doesn't It might look slightly different. I might feel a 0:31:05.400 --> 0:31:07.080 lot more blue, I might feel a lot more black, 0:31:07.120 --> 0:31:10.040 depending on how yours is constructed. But for the most part, 0:31:10.480 --> 0:31:12.160 a lot of the features are kind of where you 0:31:12.200 --> 0:31:15.760 expect them to be, and it's not at all difficult 0:31:15.760 --> 0:31:17.960 to get it up and running, which is not something 0:31:18.120 --> 0:31:19.200 against Cooper said earlier. 0:31:19.200 --> 0:31:21.640 We could say about things like PGP. 0:31:22.320 --> 0:31:25.200 Yeah, I wanted to kind of move on to talking 0:31:25.240 --> 0:31:28.040 about other apps and their security or lack of it, 0:31:28.080 --> 0:31:31.040 and I think we should start probably by talking about Telegram, 0:31:31.120 --> 0:31:33.920 because that's probably close to top of the list of 0:31:33.960 --> 0:31:38.800 things people use for secure communications that is not nearly 0:31:38.840 --> 0:31:43.160 as secure as they think. So, yeah, I wanted to 0:31:43.200 --> 0:31:45.840 kind of chat with you about like why that is, 0:31:45.920 --> 0:31:48.120 and I specifically I wanted to talk one of the 0:31:48.160 --> 0:31:52.400 things that is frustrating about Telegram is they kind of 0:31:52.480 --> 0:31:54.560 have they have like a secret chat or private chat, 0:31:54.600 --> 0:31:57.480 like they have a couple of different options that don't 0:31:57.920 --> 0:32:01.760 necessarily mean what they sound like they mean to most people. 0:32:01.840 --> 0:32:04.800 Yeah, so that's actually one thing our report found. So 0:32:04.920 --> 0:32:07.520 private chat and secret chat are in fact. 0:32:07.560 --> 0:32:08.200 The same thing. 0:32:09.240 --> 0:32:12.560 They're just called slightly different things in the app, which 0:32:12.640 --> 0:32:15.120 for for again, for those listening. 0:32:14.760 --> 0:32:17.400 That are don't have the background in design, that's bad design. 0:32:17.480 --> 0:32:21.920 That's actually not that's not professional, that's a that is 0:32:22.040 --> 0:32:25.680 a mistake. There's no reason for a feature to have 0:32:25.760 --> 0:32:29.720 like two different names inside of inside of your software. 0:32:30.760 --> 0:32:33.560 And so I don't know if that's an oversight on 0:32:33.600 --> 0:32:38.080 their part. I'm assuming so, but like those two things 0:32:38.200 --> 0:32:40.880 correlate to the same feature, and so they should actually 0:32:40.880 --> 0:32:43.719 be called the same thing. But then even further that 0:32:43.760 --> 0:32:46.720 being said, what does private mean to a user? 0:32:46.800 --> 0:32:47.960 What does secret mean? 0:32:49.440 --> 0:32:53.680 You know, Facebook Messenger they call their encrypted message secure 0:32:53.880 --> 0:32:55.040 or no, they also call it secret. 0:32:55.080 --> 0:32:57.720 Sorry, they also call it secret. But does that mean security? 0:32:57.720 --> 0:32:58.720 Does that mean encrypted? 0:32:58.880 --> 0:33:02.640 And so that's like one of the one of the 0:33:02.640 --> 0:33:05.880 weird things where it's like, you know, I think by 0:33:06.000 --> 0:33:10.280 using a very sort of like normalize or culturally almost 0:33:10.320 --> 0:33:13.840 like emotional name like private, it makes something seem like 0:33:14.120 --> 0:33:17.600 it's actually quite safe, when in fact it's not. And 0:33:17.640 --> 0:33:20.760 there's a variety of reasons as why, like Telegram is 0:33:20.800 --> 0:33:24.680 not not a very secure app that I will let Cooper. 0:33:24.440 --> 0:33:25.040 Talk about more. 0:33:25.520 --> 0:33:28.680 Yeah, I would never advise anybody to have a chat 0:33:29.080 --> 0:33:33.320 over Telegram if they are concerned about the privacy of that. Yeah, 0:33:33.360 --> 0:33:38.040 So we were talking about friction and the fact that 0:33:38.760 --> 0:33:42.600 and encrypted chats are not the default in Telegram creates 0:33:42.600 --> 0:33:47.720 a friction for users to have an actually secure chat. 0:33:47.800 --> 0:33:49.920 Right you have to go remember to turn it. 0:33:49.880 --> 0:33:53.040 On, and you can only turn it on turn it 0:33:53.080 --> 0:33:57.680 on individually per message. It's not like an overall feature 0:33:57.880 --> 0:34:01.200 on Telegram or Facebook Messenger. You have to go select 0:34:01.200 --> 0:34:06.200 a specific like the specific conversation per conversation, which is 0:34:06.880 --> 0:34:09.399 and another thing ourper gets into is how also those 0:34:09.520 --> 0:34:12.640 chats don't look very different. They look almost identical to 0:34:13.239 --> 0:34:17.239 a normal chat. So for for low vision users or 0:34:17.280 --> 0:34:21.160 anyone with any kind of like disability, especially a vision 0:34:21.239 --> 0:34:25.960 related disability, it's almost impossible to it's like nearly impossible 0:34:25.960 --> 0:34:29.759 to recognize which chat you're using if you're looking at. 0:34:29.640 --> 0:34:30.360 The chat logs. 0:34:31.080 --> 0:34:34.680 Yeah, outside of that, like if people, you know, in 0:34:34.760 --> 0:34:37.080 terms of like things, that may not be options right now, 0:34:37.120 --> 0:34:40.840 I think basically everyone listening signal is a perfectly viable option. 0:34:40.880 --> 0:34:43.920 But it's not impossible that, for example, you might wind 0:34:44.000 --> 0:34:46.640 up in a country where, even if there's not a 0:34:46.680 --> 0:34:49.880 specific law against it, there is a precedent established that 0:34:49.920 --> 0:34:52.279 if you have signal on your phone, you know, it 0:34:52.719 --> 0:34:55.600 can be at least used as a justification for charges 0:34:55.640 --> 0:34:58.200 that you are planning to use. Like you know, with Atlanta, 0:34:58.280 --> 0:35:01.440 people are getting charges because they had a lawyer's name 0:35:01.480 --> 0:35:04.440 written on their arm right, and and so the state saying, well, 0:35:04.480 --> 0:35:07.440 that's evidence that we're planning to commit a crime. You know, 0:35:07.640 --> 0:35:09.880 that doesn't mean that convictions will go through in that 0:35:09.960 --> 0:35:11.480 kind of thing, but it may be a reason why 0:35:11.560 --> 0:35:15.000 signal might not be an option, or say, you know, 0:35:15.080 --> 0:35:17.520 something comes out about it that makes it seem less secure. 0:35:18.000 --> 0:35:22.360 What are other good or or acceptable options? And I 0:35:22.400 --> 0:35:24.720 know when we're talking about this, these are often options 0:35:24.719 --> 0:35:27.319 that require more input and work from the user in 0:35:27.440 --> 0:35:30.120 order to maximize their potential security. But I do think 0:35:30.160 --> 0:35:31.600 it's good to like let people kind of know what 0:35:31.640 --> 0:35:32.439 else is out there. 0:35:32.840 --> 0:35:36.920 Yeah, so when signal isn't an option, WhatsApp is actually 0:35:37.040 --> 0:35:41.439 not a bad option. So WhatsApp it is owned by Meta, 0:35:41.640 --> 0:35:44.400 which is a you know which is which is you 0:35:44.440 --> 0:35:48.880 know not which is not ideal? But WhatsApp actually uses 0:35:48.960 --> 0:35:53.399 the same encryption protocol as signal. Uh so, like under 0:35:53.400 --> 0:35:55.000 the hood, the way that the you know, the way 0:35:55.040 --> 0:35:57.000 that the math works to hide your messages from the 0:35:57.120 --> 0:36:02.400 NSA is exactly the same, right, and they've implemented it well. 0:36:02.440 --> 0:36:05.759 You know, there are a few more steps that, you know, 0:36:05.800 --> 0:36:07.960 a few more precautions that you need to take with WhatsApp, 0:36:08.040 --> 0:36:11.120 like making sure that your chats aren't backed up being 0:36:11.120 --> 0:36:16.399 the main one. But WhatsApp is certainly good enough, right 0:36:16.440 --> 0:36:20.760 if you're if you're you know, chat networks aren't using signal, 0:36:20.840 --> 0:36:23.000 if you're in a country where you can't use signal, right, 0:36:23.120 --> 0:36:27.680 Like WhatsApp has two billion users, I'm you know, it's 0:36:27.719 --> 0:36:30.040 it's you can use WhatsApp almost anywhere in the world. 0:36:30.080 --> 0:36:32.319 It's and it's ubiquitous enough that it's not going to 0:36:32.440 --> 0:36:35.120 mark you as you know, somebody with something to hide, right, 0:36:35.680 --> 0:36:37.680 And like, and I don't want to I don't want 0:36:37.719 --> 0:36:42.000 to discount what's app. Right, getting two billion people to 0:36:42.520 --> 0:36:46.840 have end to end encrypted messaging by default overnight basically 0:36:47.360 --> 0:36:52.080 was a major cupe. Like that that was world changing, right, 0:36:52.320 --> 0:36:57.280 and like they they really do deserve applause for that obviously, 0:36:57.520 --> 0:36:59.720 you know, I think partly because of their scale, partly 0:36:59.719 --> 0:37:02.520 because they're own b meta, right, they haven't taken all 0:37:02.560 --> 0:37:06.880 of these same steps, Like they do have more metadata 0:37:07.320 --> 0:37:11.239 on their servers than Signal does. Right. But if that's 0:37:11.239 --> 0:37:13.080 your option, that is a fine option. 0:37:13.880 --> 0:37:17.560 Yeah, I think that's really good to know, particularly since 0:37:18.040 --> 0:37:21.960 options are always more secure than not having any kind 0:37:21.960 --> 0:37:24.280 of a backup plan totally. 0:37:24.719 --> 0:37:28.000 And if people are like even slightly nervous about WhatsApp, 0:37:28.200 --> 0:37:31.480 of great things, they do have disappearing messages. The downside 0:37:31.560 --> 0:37:35.000 is like the fastest disappearing message is only twenty four hours. 0:37:35.000 --> 0:37:39.040 But that's something that again you still have, and that's 0:37:39.280 --> 0:37:41.520 like that is that is an amazing feature. 0:37:42.280 --> 0:37:45.640 Yeah, and that kind of gets into also what kind 0:37:45.680 --> 0:37:48.960 of stuff you can do in order to maximize the 0:37:49.040 --> 0:37:52.200 value of features like that. Like, for example, if you're 0:37:52.480 --> 0:37:55.560 coming back into the country or a country and your 0:37:55.600 --> 0:37:59.920 phone gets confiscated by customs or whatever because security so 0:38:00.160 --> 0:38:02.600 uses have some sort of eye on you for whatever reason. 0:38:03.000 --> 0:38:07.880 If you've got you know, thumbprint log in or face 0:38:07.920 --> 0:38:09.920 log in, they're going to get into that phone right 0:38:10.320 --> 0:38:12.799 in your twenty four hour delete thing may not have 0:38:12.840 --> 0:38:15.799 gotten taken care of everything. If you've got like a 0:38:15.840 --> 0:38:20.640 complicated eight digit password and no biometrics enabled, maybe depending 0:38:20.680 --> 0:38:23.040 on where you are and whatnot, that'll keep your phone 0:38:23.080 --> 0:38:25.680 locked long enough for those messages to get deleted, right, Like, 0:38:25.719 --> 0:38:28.680 it's all about kind of maximizing the chances that something. 0:38:28.400 --> 0:38:32.680 Like that helps. Yeah, exactly. We definitely recommend that people 0:38:32.719 --> 0:38:35.920 turn on disappearing messages. I think that that's just a 0:38:35.960 --> 0:38:40.600 good sensible default to have. Also definitely recommend that if 0:38:40.640 --> 0:38:43.480 you're going to be in a situation where you think 0:38:43.480 --> 0:38:45.400 you're going to be, you know, there's a higher likelihood 0:38:45.400 --> 0:38:47.960 if you interacting with law enforcement, if you're crossing a border, 0:38:48.000 --> 0:38:50.759 if you're going to a protest turn off the biometric 0:38:50.960 --> 0:38:54.600 unlock on your phone. Certainly, especially in the US, there's 0:38:54.719 --> 0:38:57.319 the case law isn't settled, but there's a lot of 0:38:57.360 --> 0:39:01.279 state courts that have decided that police can force you 0:39:01.320 --> 0:39:03.560 to unlock your phone with your biometrics and that that's 0:39:03.600 --> 0:39:06.759 totally fine. So this, you know, in the in the 0:39:06.840 --> 0:39:09.120 US context, is a good idea in any context. I 0:39:09.120 --> 0:39:11.040 think it's a good idea if you're at heightened risk 0:39:11.120 --> 0:39:12.680 to turn off totally. 0:39:14.080 --> 0:39:15.560 I mean, one thing we're also a big fan of 0:39:15.719 --> 0:39:18.000 is figuring out too like and this is again where 0:39:18.000 --> 0:39:20.759 threat modeling is so key, is like, is this a 0:39:20.760 --> 0:39:23.759 circumstance where you need your phone or another thing that 0:39:24.400 --> 0:39:26.040 you know you can always do if you are nervous 0:39:26.080 --> 0:39:29.800 about traveling across the border, is you can delete signal 0:39:29.840 --> 0:39:33.240 and reinstall it and everything is gone. You can delete 0:39:33.239 --> 0:39:36.560 WhatsApp temporarily while you're crossing a border so it's not 0:39:37.200 --> 0:39:40.600 on your phone. You know, there are things like that 0:39:40.640 --> 0:39:43.720 you can do if you feel comfortable wiping your phone, 0:39:43.760 --> 0:39:47.680 that's something also you can do. You know, these are 0:39:47.680 --> 0:39:50.560 all again these are these are these are different things, 0:39:50.600 --> 0:39:52.240 and I think this is one of the things our 0:39:52.360 --> 0:39:55.279 report I don't remember how too much we get into 0:39:55.280 --> 0:39:57.080 a bit something that at least we've been thinking about. 0:39:57.080 --> 0:40:00.040 Cooper and I run a little lab called Complication, and 0:40:01.400 --> 0:40:04.440 one of the things we've been thinking about there is 0:40:04.520 --> 0:40:07.240 just also how do we instill sort of like better, 0:40:07.440 --> 0:40:10.680 better holistic practices where we understand that a phone is 0:40:10.760 --> 0:40:15.320 just one component of our safety and so like secure messaging, 0:40:15.360 --> 0:40:17.920 encrypted messaging is one component of that safe safety. So 0:40:18.000 --> 0:40:20.279 like what are other things we can do? 0:40:20.760 --> 0:40:21.840 And some of that can. 0:40:21.640 --> 0:40:24.680 Be you know, wiping your phone if traveling, if that 0:40:24.719 --> 0:40:26.560 makes sense for you, or if that's a thing that 0:40:26.640 --> 0:40:30.760 makes you feel safer, or removing certain apps and then 0:40:31.160 --> 0:40:33.560 you know, reinstalling them, reinstalling them later. 0:40:34.280 --> 0:40:37.080 Yeah, yeah, and it and it really is holistic. Right, 0:40:37.160 --> 0:40:39.640 Like a thing that a thing that people need to 0:40:39.680 --> 0:40:43.000 keep in mind is that, you know, disappearing messages can't 0:40:43.040 --> 0:40:50.760 stop an untrustworthy uh conversation partner, right Like if if 0:40:50.880 --> 0:40:53.800 my conversation partner is untrustworthy, they can take screenshots of 0:40:53.840 --> 0:40:57.160 the messages, right, they can you know, go they can 0:40:57.200 --> 0:41:01.080 go snitch to law enforcement about what I've hold them. Right, 0:41:02.480 --> 0:41:06.200 Encrypted messaging, discipree messages, These are not panaspeas. Right, you 0:41:06.280 --> 0:41:09.640 still have to you still have to keep all of 0:41:09.680 --> 0:41:17.160 your other aspects of security as well, right, So don't 0:41:17.320 --> 0:41:19.680 entirely rely on these technologies to save you, right, you 0:41:19.719 --> 0:41:21.680 have to also trust the people you're working with and 0:41:22.040 --> 0:41:23.440 build these layers of security. 0:41:23.480 --> 0:41:26.000 Yet it's true, I mean, Cooper, you could leak all 0:41:26.000 --> 0:41:29.000 of my secrets right now on this podcast and them. 0:41:28.920 --> 0:41:30.120 That too, what a gentleman. 0:41:30.719 --> 0:41:33.600 And that is that is the other thing, right where 0:41:34.640 --> 0:41:37.879 when it comes to like what is secure, one thing 0:41:37.920 --> 0:41:40.520 to remember is that signal for all the good things 0:41:40.560 --> 0:41:43.120 about it, Nothing, nothing at all about that app stops 0:41:43.520 --> 0:41:45.719 the recipient of a message from you from taking a 0:41:45.760 --> 0:41:49.520 screen grab or just handing their phone over to their 0:41:49.520 --> 0:41:54.160 friendly local federal agent, right, which is always you know, 0:41:54.960 --> 0:41:56.560 we don't want to be I'm not trying to be 0:41:56.640 --> 0:41:58.960 a security nihilist here. I think you know, there's no 0:41:59.120 --> 0:42:04.880 replacing communication over phones in many situations. But if you are, 0:42:04.960 --> 0:42:11.040 for example, going to be transferring a bunch of Plan 0:42:11.160 --> 0:42:15.279 B pills in an area where that is prosecutable, that 0:42:15.320 --> 0:42:17.799 probably shouldn't go on your phone in that language. Right, 0:42:17.880 --> 0:42:20.600 Perhaps you know you could come up with a clever 0:42:20.719 --> 0:42:26.400 codeword or whatever. But don't you know it is security 0:42:26.480 --> 0:42:28.640 is like you said, holistic. You know you should not 0:42:28.800 --> 0:42:31.560 be looking at it as just like, well the app 0:42:31.640 --> 0:42:32.120 is secure. 0:42:32.160 --> 0:42:32.879 So that's enough. 0:42:33.760 --> 0:42:35.400 I mean, one thing I also want people sort of 0:42:35.400 --> 0:42:37.360 think about too, because that's a really great point Robert, 0:42:37.520 --> 0:42:39.799 is like, we do all different kinds of things every 0:42:39.840 --> 0:42:43.320 day in our lives that could, you know, in dangerous. 0:42:43.360 --> 0:42:45.600 Like I think a lot of the work I do 0:42:45.719 --> 0:42:47.360 is I work a lot with people facing all different 0:42:47.400 --> 0:42:50.520 kinds of online harassment. So like falling in love, for example, 0:42:50.600 --> 0:42:52.640 is a dangerous thing to do. You could have your 0:42:52.680 --> 0:42:56.799 heart broken or that person could hurt you. Learning how 0:42:56.800 --> 0:43:01.640 to trust people, you know, crossing the street, deciding to 0:43:01.760 --> 0:43:04.440 jaywalk right, all different things we do sort of. 0:43:04.360 --> 0:43:06.520 Every day actually can expose us to harm. 0:43:06.560 --> 0:43:08.400 And so one thing I think for people listening to 0:43:08.440 --> 0:43:11.040 keep in mind is that's the same one we have conversations. 0:43:11.120 --> 0:43:14.200 And I think a way to avoid nihilism is just 0:43:14.239 --> 0:43:17.120 to remember that that every day we are sort of 0:43:17.120 --> 0:43:19.920 going out there and actually being incredibly brave just by 0:43:19.960 --> 0:43:23.480 living our everyday lives, by deciding to be in community 0:43:23.520 --> 0:43:28.200 and have friendships and have relationships, and in my case, 0:43:28.200 --> 0:43:31.680 I love jaywalking, and no one around me does, and 0:43:32.040 --> 0:43:35.920 that's why that's my choice. And I have not yet 0:43:36.239 --> 0:43:38.080 gotten hit by a car jaywalking. 0:43:38.880 --> 0:43:42.080 I think it's good to look at this the same way. 0:43:42.120 --> 0:43:46.719 There's a concept that the military has sort of developed 0:43:46.719 --> 0:43:49.279 when talking about how not to die when you're in 0:43:49.320 --> 0:43:52.359 a gunfight or something. It's called the survivability onion, right, 0:43:52.640 --> 0:43:55.000 And I think it's extremely useful both if you're talking 0:43:55.040 --> 0:43:56.759 about like, well, I'm going to a protest and there 0:43:56.800 --> 0:43:58.919 will be violence there, you know, should I wear armor, 0:43:58.920 --> 0:44:01.880 et cetera. But it's also just really it's really useful 0:44:01.880 --> 0:44:05.160 with any kind of security, and and the onion, it's 0:44:05.200 --> 0:44:08.040 it's envision, doesn't on you because like the largest outside 0:44:08.120 --> 0:44:10.920 chunk of it is don't be seen, don't be acquired, 0:44:10.960 --> 0:44:13.120 which means somebody actually getting you in their head sights. 0:44:13.640 --> 0:44:17.960 Don't be a hit, which means being behind cover or something. 0:44:18.200 --> 0:44:20.640 And then the very internal part of it is like, 0:44:20.719 --> 0:44:22.800 have some sort of armor in case you are shot. 0:44:23.160 --> 0:44:24.040 But if you if the. 0:44:24.080 --> 0:44:27.640 Armor is useful, the majority of the onion has already failed. 0:44:27.760 --> 0:44:27.920 Right. 0:44:28.320 --> 0:44:32.480 If encryption is useful, that is not a dissimilar sort 0:44:32.520 --> 0:44:35.040 of situation. Right, So there's a there's a degree of 0:44:35.840 --> 0:44:41.000 canniness is super helpful and thinking about like what is 0:44:41.320 --> 0:44:44.799 what is visible about me? If I'm doing something, I 0:44:44.880 --> 0:44:47.480 know that I have to be extra concerned about the 0:44:47.520 --> 0:44:51.920 state seeing what is visible about me from the outside, 0:44:52.200 --> 0:44:53.399 you know, tobly, I mean. 0:44:53.360 --> 0:44:55.040 I think that's an amazing thing to think about. Like 0:44:55.480 --> 0:44:57.920 where where are you sending a text message? Are you 0:44:58.000 --> 0:45:00.879 in a place in which like someone can lean over, 0:45:00.960 --> 0:45:04.680 Like I'm the nosiest motherfucker, and all the time I'm 0:45:04.760 --> 0:45:07.400 constantly like like looking around being like what's that person 0:45:07.440 --> 0:45:10.240 watching on an airplane? Or like if someone is sitting 0:45:10.239 --> 0:45:12.960 next to me scrolling, So like you wouldn't want to 0:45:12.960 --> 0:45:15.360 like send a sensitive text message like next to me, 0:45:15.400 --> 0:45:18.520 because I'd be like, that's that's interesting fodder. 0:45:18.600 --> 0:45:23.200 That's kind of a show Texas to Cooper later, you know. 0:45:23.239 --> 0:45:24.919 And so I think it's important to think about that, 0:45:25.200 --> 0:45:28.200 Like who's around you? Is this is like how are 0:45:28.280 --> 0:45:31.400 you describing something? Do you know the person you're messaging? 0:45:31.440 --> 0:45:33.560 If you're in a group message, you know everybody there? 0:45:33.600 --> 0:45:35.759 Like do you trust all of them? 0:45:36.760 --> 0:45:37.000 You know? 0:45:37.040 --> 0:45:39.200 And if you're ever nervous there are this is I 0:45:39.200 --> 0:45:42.279 guess the upside also to in person conversations. You can have, 0:45:43.200 --> 0:45:45.440 you know, a phone call or an in person conversation 0:45:45.560 --> 0:45:48.640 with someone. Right if you're really not sure or you 0:45:48.680 --> 0:45:51.839 don't feel comfortable even sending something over signal, that might 0:45:51.840 --> 0:45:53.000 be the time to be like, hey, do you want 0:45:53.000 --> 0:45:55.000 to meet up and get a coffee and then you know, 0:45:56.719 --> 0:46:01.160 try to find a discreet place to have have a conversation. 0:46:02.000 --> 0:46:07.680 Yeah, yeah, I do want to roll to ads real quick. 0:46:07.719 --> 0:46:09.799 One second, and I think Cooper had something to say, 0:46:09.840 --> 0:46:25.880 and we'll we'll continue, but first products ah, we're back Cooper. 0:46:26.200 --> 0:46:28.440 You look like you had something to add. 0:46:28.200 --> 0:46:31.080 On that, nothing particularly serious, just that. I think that 0:46:31.080 --> 0:46:32.920 that's I think that that's really good advice for the 0:46:32.920 --> 0:46:36.080 military and absolutely justifies the nine hundred billion dollars. 0:46:39.040 --> 0:46:42.799 Yeah, I'm glad they put together a fucking graphic. I 0:46:42.840 --> 0:46:45.879 wonder how many billions of dollars that did best I could. 0:46:45.920 --> 0:46:49.360 I could make a graphic for hundreds of millions of dollars. 0:46:49.600 --> 0:46:52.160 Yeah, if anybody, if anybody wants to fund us for 0:46:52.239 --> 0:46:54.319 hundreds of millions, we will will do it less now 0:46:54.640 --> 0:46:56.239 a year, hundreds of millions. 0:46:56.320 --> 0:46:59.000 We have so many good T shirt ideas and sticker 0:46:59.080 --> 0:47:03.239 ideas y'all like so many good ones, so many unhinged 0:47:03.239 --> 0:47:04.880 ones that the world needs to see. 0:47:05.280 --> 0:47:08.759 Yeah, I mean I do. I guess just because of 0:47:08.800 --> 0:47:10.920 the amount of time I've spent thinking about this stuff 0:47:11.200 --> 0:47:14.080 from my old job. There are a couple of concepts 0:47:14.200 --> 0:47:18.920 from military planning I think about in this context, and 0:47:18.960 --> 0:47:21.520 one of them that I also think is relevant to 0:47:21.520 --> 0:47:23.800 what we're talking about with friction is the concept of 0:47:23.840 --> 0:47:26.480 an ode loop, right, which is how do you win 0:47:27.360 --> 0:47:30.480 and combat against an opponent, And it's by disrupting this 0:47:30.520 --> 0:47:32.200 thing called the ode loop. And the ODO loop is 0:47:32.239 --> 0:47:37.719 how an adversary carries out actions in a conflict like this, right, 0:47:38.120 --> 0:47:42.360 and the steps you have to go for observe, orient, decide, 0:47:42.400 --> 0:47:44.799 and act. And if you can disrupt any stage of that, 0:47:45.120 --> 0:47:48.280 you can stop them from taking actions, right, which stops 0:47:48.320 --> 0:47:52.480 them from being able to harm you. And the good 0:47:52.600 --> 0:47:56.960 security is going to impact all three of those things, right, 0:47:57.000 --> 0:47:58.719 It's going to stop them from being able to see 0:47:58.760 --> 0:48:02.200 you sometimes if they can see you stuff like you know, 0:48:02.560 --> 0:48:04.840 you were just talking, we were just talking earlier about 0:48:05.600 --> 0:48:07.760 link previews, right, and how that can kind of expose 0:48:07.880 --> 0:48:11.319 maybe who you're in communication with potentially, well, that could 0:48:11.320 --> 0:48:13.719 allow the state to orient themselves to you and to 0:48:13.840 --> 0:48:17.879 your friends, right, And obviously stuff like locking down your 0:48:17.880 --> 0:48:21.800 devices not having unnecessarily info online can stop them bring 0:48:22.080 --> 0:48:24.320 being able to decide, you know, what you're doing and 0:48:24.640 --> 0:48:27.480 how they should respond to that. And I think that's 0:48:27.520 --> 0:48:30.480 also good if you're thinking, if you're not just somebody 0:48:30.480 --> 0:48:32.839 who is concerned about your security like most people are, 0:48:32.880 --> 0:48:35.720 because it's good to have some security. If you're actually 0:48:36.400 --> 0:48:40.120 dealing with the state or a corporation as an adversary 0:48:40.200 --> 0:48:43.400 in some way, it can be useful to think about 0:48:43.520 --> 0:48:45.440 your security culture in those terms. 0:48:46.560 --> 0:48:50.279 Yeah. Absolutely, I think that's absolutely right. It's it's and 0:48:50.320 --> 0:48:52.040 I think that it's you know, it points to you 0:48:52.160 --> 0:48:56.719 like we should, we should understand what the you know, 0:48:57.080 --> 0:49:00.480 mode of thinking of our adversaries is, right, like we 0:49:00.640 --> 0:49:04.359 you know, we should if your adversary is the NSA, right, 0:49:04.360 --> 0:49:08.160 which is like probably actually not most people in the US, 0:49:08.200 --> 0:49:10.960 Like for most US activists, the NSA is not actually 0:49:11.040 --> 0:49:14.279 your biggest adversary, right, Like your biggest adversary is going 0:49:14.360 --> 0:49:17.000 to be local police, right, your biggest adversary is going 0:49:17.080 --> 0:49:20.080 to be you know, the the you know, somebody like 0:49:20.120 --> 0:49:22.480 your abusive partner, right, and you need to. And this 0:49:22.520 --> 0:49:27.440 is why threat modeling is important, because you need to 0:49:27.440 --> 0:49:29.919 to really to really think about, you know, think through 0:49:30.040 --> 0:49:32.279 like you know, well, okay, wait, am I actually worried 0:49:32.280 --> 0:49:34.279 about protecting myself from the NSA? Or am I more 0:49:34.320 --> 0:49:38.120 worried about uh uh, you know the the racist police 0:49:38.120 --> 0:49:41.000 officer that drives down my street every day? Right? And yeah, 0:49:41.040 --> 0:49:43.440 probably it's the latter. And so you can you can 0:49:43.480 --> 0:49:47.279 take a lot more useful actions, right uh. And and 0:49:47.360 --> 0:49:51.239 you know you can you can you know, break that 0:49:51.360 --> 0:49:55.600 oda loop for him once you know actually what it is. Right, Yeah, 0:49:55.680 --> 0:49:58.000 if you're defending yourself against the NSA, you're gonna leave 0:49:58.000 --> 0:50:01.879 yourself wide open to the actual threat. Yeah. 0:50:01.880 --> 0:50:05.200 It's totally I think a great example. And I don't 0:50:05.239 --> 0:50:07.680 mean to be like quote unquote sub tweeting somebody here, 0:50:07.719 --> 0:50:10.640 but I've known a couple of folks like this. It's like, 0:50:10.640 --> 0:50:12.920 if you have if you're super paranoid, you're not putting 0:50:12.920 --> 0:50:16.000 anything online, You're only talking with your close friends, you 0:50:16.120 --> 0:50:18.799 use a dumb phone, you have burners, but you also 0:50:18.880 --> 0:50:20.560 drive around with a shitload of weed in your car 0:50:20.600 --> 0:50:22.839 in a state where that's illegal. It's like, well, like 0:50:23.080 --> 0:50:27.239 your threat modeling is not great in that situation, right, 0:50:27.360 --> 0:50:29.080 or like I do all that, but I carry in 0:50:29.120 --> 0:50:32.319 a legal handgun with me wherever I go. It's like, well, 0:50:32.400 --> 0:50:34.880 it may be more of a threat than your phone. 0:50:35.120 --> 0:50:36.840 My partner the other day was like, what if I 0:50:36.840 --> 0:50:38.120 got a dumb phone? I was like, what if I 0:50:38.160 --> 0:50:41.759 divorced you? Like like what if? 0:50:43.239 --> 0:50:44.120 They were like what do you mean? 0:50:44.160 --> 0:50:45.640 And I was like, well, I'm going to be the 0:50:45.640 --> 0:50:48.840 one using all the maps for both of us, yeah, 0:50:48.880 --> 0:50:52.120 and having to google all the dumb shit you want 0:50:52.120 --> 0:50:55.000 at Google. That doesn't make I'm now your weakest link, 0:50:55.040 --> 0:50:57.399 like go fuck yourself. But also I was like, I'm 0:50:57.440 --> 0:51:02.560 absolutely not going to be your your your Google maps bitch, 0:51:02.640 --> 0:51:06.239 Like I'm not not doing that. But but I mean 0:51:06.280 --> 0:51:08.560 I think also, you know, to both of y'all's points 0:51:08.560 --> 0:51:10.560 to get serious again for a second. I mean, you know, 0:51:10.600 --> 0:51:15.040 like my threat model, for example, might be similar or 0:51:15.040 --> 0:51:18.680 slightly different, maybe slightly less serious than than Cooper's. But 0:51:18.840 --> 0:51:21.520 you know, like some of the like the the the 0:51:21.600 --> 0:51:23.560 journalists in India we were working with, have quite a 0:51:23.640 --> 0:51:26.840 high threat model, right, Like, yeah, the Indian police force 0:51:27.200 --> 0:51:30.920 are very much like the NSA. They're very talented they 0:51:30.960 --> 0:51:33.200 have a lot of money and tech at their disposal, 0:51:33.560 --> 0:51:36.840 and that might be different for some of the activists 0:51:36.880 --> 0:51:39.480 we're working with, let's say in like Louisiana or Texas, right, 0:51:40.719 --> 0:51:44.160 but the differences is, like we're still talking about I 0:51:44.160 --> 0:51:48.920 would argue two brutal police forces that just have different 0:51:49.000 --> 0:51:51.839 means of disposal at their hands. So like the Louisiana 0:51:52.080 --> 0:51:55.880 police are our groups you should totally be worried about. 0:51:56.120 --> 0:51:58.880 They might not be able to hack your phone, but 0:51:59.400 --> 0:52:00.600 maybe eventually they could. 0:52:00.960 --> 0:52:03.000 But there are other there are obviously other things that 0:52:03.040 --> 0:52:03.719 were about them. 0:52:03.760 --> 0:52:07.000 But you know, in the context of like some of 0:52:07.040 --> 0:52:09.800 the folks who are working with in the South, like 0:52:09.840 --> 0:52:13.160 reproductive justice activists, some of the things are probably much 0:52:13.200 --> 0:52:17.160 more serious in terms of your threat model would be 0:52:17.200 --> 0:52:20.680 like a nurse for someone who let's say is miscarring 0:52:20.960 --> 0:52:23.359 or has sought an abortion. And this is something Kate 0:52:23.360 --> 0:52:26.719 Bertosh from the Digital Defense Fund, a friend of of 0:52:26.880 --> 0:52:29.720 you know ours, has talked about where like the people 0:52:29.800 --> 0:52:31.799 that are supposed to take care of you might be 0:52:31.840 --> 0:52:35.359 the ones that are actually your your biggest threat, right, 0:52:35.640 --> 0:52:37.960 the ones that have heard you say something or you've 0:52:37.960 --> 0:52:41.200 can fight it in for example, and that is kind 0:52:41.239 --> 0:52:42.640 of a horrifying thing to think about. 0:52:42.680 --> 0:52:44.399 But that is, that is a thing you. 0:52:44.400 --> 0:52:47.479 Have to threat model, right, is is can I trust 0:52:47.520 --> 0:52:49.640 this person? How am I describing? 0:52:49.760 --> 0:52:50.760 You know? What's happening? 0:52:51.400 --> 0:52:51.640 Yeah? 0:52:52.200 --> 0:52:56.239 Yeah, absolutely, Well, did y'all have anything else you wanted 0:52:56.280 --> 0:52:59.080 to make sure to get into in this conversation? There's 0:52:59.080 --> 0:53:01.360 so much more in your in the great paper you 0:53:01.719 --> 0:53:05.200 helped co author, What is Secure and Analysis of Popular 0:53:05.239 --> 0:53:08.600 Messaging Apps on the Tech Policy Press. But yeah, is 0:53:08.600 --> 0:53:10.560 there anything else y'all wanted to really make sure you 0:53:10.640 --> 0:53:11.640 hit before we roll out? 0:53:12.160 --> 0:53:14.400 Yeah? Please don't use telegram for a variety of reasons, 0:53:14.440 --> 0:53:17.560 but also like it's very unclear how they respond to 0:53:17.600 --> 0:53:20.200 any law enforcement or government. They don't say anything, and 0:53:20.200 --> 0:53:22.520 it's kind of impossible to reach anyone that works there. 0:53:23.160 --> 0:53:27.000 Please don't use Facebook Messenger other than maybe sending memes. 0:53:28.520 --> 0:53:31.160 There's a lot of really gross surveillance capitalism inside a 0:53:31.200 --> 0:53:35.160 Facebook messenger that the paper gets into. But effectively, Meta 0:53:35.239 --> 0:53:38.720 is building this weird, sprawling infrastructure inside a Facebook Messenger 0:53:38.719 --> 0:53:40.640 and try to link Facebook and Instagram. 0:53:41.080 --> 0:53:43.239 And one of the things we noticed is. 0:53:43.200 --> 0:53:45.919 That, like if you've blocked someone on Instagram or mute 0:53:45.920 --> 0:53:48.799 to them, but you haven't blocked remuted them on Facebook, 0:53:49.160 --> 0:53:51.680 that your stories, like all those stories are still coming 0:53:51.719 --> 0:53:55.560 across in messengers, so you can still see content from 0:53:55.600 --> 0:53:59.480 someone because it's linking both of those both of those profiles. 0:54:00.239 --> 0:54:02.719 So you know, you could see how partaking like an 0:54:02.760 --> 0:54:05.720 online harassment lens like why that's why that's really bad, 0:54:06.320 --> 0:54:11.440 that's really harmful and could be potentially you know, upsetting 0:54:11.480 --> 0:54:13.000 and triggering for folks. 0:54:14.640 --> 0:54:18.040 Yeah, I'll add that. I think my the major thing 0:54:18.080 --> 0:54:21.320 I want people to to think about is that encryption 0:54:21.440 --> 0:54:24.239 really does work, and it works really well. And we 0:54:24.280 --> 0:54:26.759 can see that because a lot of countries right now 0:54:27.120 --> 0:54:30.160 are trying to pass laws that either weaken or byan 0:54:30.320 --> 0:54:34.160 encryption right and in fact, the UK uh did passed, 0:54:34.480 --> 0:54:37.399 did just pass such a law, the online Safety built 0:54:37.400 --> 0:54:40.680 in the UK. And so it's really important that we 0:54:41.480 --> 0:54:43.799 that we you know, push back against these laws and 0:54:43.840 --> 0:54:48.160 fight back against these laws and and whatever we can, right. 0:54:48.200 --> 0:54:49.720 And I'm not I'm not coming at. 0:54:49.560 --> 0:54:53.799 This as somebody who's a big believer in the you know, 0:54:53.880 --> 0:54:57.040 in in incrementalism and in working with governments, but I 0:54:57.120 --> 0:54:59.640 think that I still think that it's really important to 0:55:00.480 --> 0:55:00.839 you know. 0:55:01.719 --> 0:55:04.040 Educate folks and push back against these laws and try 0:55:04.040 --> 0:55:08.040 to not let these pasts because these will be really 0:55:08.080 --> 0:55:10.200 bad for all of us totally. 0:55:10.400 --> 0:55:12.680 And not to defend the Online Safet Bill, because I 0:55:12.680 --> 0:55:14.719 would never do that. I'll go to my grave not 0:55:15.280 --> 0:55:19.000 speaking highly of it, only speaking critically at least, like 0:55:19.080 --> 0:55:24.280 the pushback from encryption experts and encryption supporters like Merrit Whitaker, 0:55:24.719 --> 0:55:28.919 president of Signal, did lead to lawmakers in the UK, 0:55:29.080 --> 0:55:33.200 for example, admitting that there's no sort of feasible safe 0:55:33.239 --> 0:55:35.359 way to build a back door, right, And that is 0:55:35.440 --> 0:55:39.719 I think also a win because of so much pushback, 0:55:39.800 --> 0:55:42.839 because of so much research, because of so much criticism 0:55:42.920 --> 0:55:46.160 that security and privacy folks gave people that are pro 0:55:46.280 --> 0:55:49.920 encryption like that, we you know, we were able to 0:55:50.000 --> 0:55:53.279 walk back that part. And I do think that's a 0:55:53.320 --> 0:55:58.439 big deal, even if there are other issues with that bill, 0:55:58.440 --> 0:56:01.480 because I think it also sends a sick pun intended 0:56:01.800 --> 0:56:07.040 to other governments as well, and I think that that's 0:56:07.040 --> 0:56:09.319 incredibly important. But yeah, I would also say just just 0:56:09.400 --> 0:56:10.479 use Signal whenever you can. 0:56:13.000 --> 0:56:17.399 But yeah, yeah, well all right, folks, that is going 0:56:17.480 --> 0:56:22.120 to be it for us here at it could happen here. Yeah, 0:56:22.160 --> 0:56:25.000 thank you all for listening, and thank you Cooper and 0:56:25.200 --> 0:56:26.680 Carolyn for coming on. 0:56:27.880 --> 0:56:29.440 Thank you for having us, yeah. 0:56:29.280 --> 0:56:30.200 And thank you for having us. 0:56:30.400 --> 0:56:34.239 You can find us on social media for now, I 0:56:34.239 --> 0:56:35.759 guess until it all. 0:56:35.600 --> 0:56:36.320 Lights on fire. 0:56:36.520 --> 0:56:38.360 Yeah, whichever one you want to trust. 0:56:40.080 --> 0:56:44.400 I'm Cooper Cue on most social media's Blue Sky Mastered 0:56:44.440 --> 0:56:45.480 on Shitter. 0:56:46.520 --> 0:56:49.799 Yeah, I'm Caroline Cinders. My first name, last name. Our 0:56:49.880 --> 0:56:54.800 lab is Convocation Research and Design Record Labs on Twitter 0:56:54.840 --> 0:56:55.360 at the moment. 0:56:55.680 --> 0:56:57.879 Hopefully we'll get be getting on Blue Sky very soon. 0:56:58.920 --> 0:57:01.560 Yeah. Yeah, probably get back on there more. 0:57:01.600 --> 0:57:01.799 Now. 0:57:01.880 --> 0:57:06.879 Twitter has gotten remarkably worse, which you know, we had 0:57:06.880 --> 0:57:11.440 a back in the day on the old something Awful forums. 0:57:11.480 --> 0:57:14.320 There was a thread in one of the debate forums 0:57:14.360 --> 0:57:17.919 about this very right wing site called Free Republic, which 0:57:17.960 --> 0:57:20.440 is like one of the earliest reservoirs of what became 0:57:20.960 --> 0:57:24.320 trump Ism, and the tagline for the thread just kind 0:57:24.360 --> 0:57:26.920 of watching these people, was there is always more and 0:57:26.960 --> 0:57:30.440 it is always worse, And boy, goddamn, if that hasn't 0:57:30.480 --> 0:57:34.720 been a continually accurate statement about the whole of social media. 0:57:34.800 --> 0:57:37.640 Right now, isn't a time amazing to watch someone just 0:57:37.720 --> 0:57:39.600 light forty billion dollars on fire. 0:57:39.800 --> 0:57:43.680 Yeah, just like yeah, totally there to it. Yeah, it's 0:57:43.720 --> 0:57:47.160 like the nihilist and me being like, wow, Comrade Musk 0:57:47.200 --> 0:57:50.040 really really taking some hits to capitalism here. 0:57:54.600 --> 0:57:56.960 It could Happen here as a production of cool Zone Media. 0:57:57.200 --> 0:57:59.880 For more podcasts from cool Zone Media, visit our website 0:58:00.000 --> 0:58:02.120 fo zonemedia dot com or check us out on the 0:58:02.160 --> 0:58:05.680 iHeartRadio app, Apple Podcasts, or wherever you listen to podcasts. 0:58:06.120 --> 0:58:08.280 You can find sources for It Could Happen Here, updated 0:58:08.320 --> 0:58:12.360 monthly at coolzonemedia dot com slash sources. Thanks for listening,