[00:00.000 --> 00:02.240]  It's been a great pleasure to announce a really awesome team of people.
[00:02.360 --> 00:05.160]  I love it when a talk comes down the speaker list and you're like, Yes!
[00:05.340 --> 00:05.840]  That talk!
[00:06.040 --> 00:06.420]  That's great!
[00:06.760 --> 00:12.840]  You're looking at the talks and I was like, two ladies from organizations that I think are just wonderful, doing amazing work.
[00:13.140 --> 00:16.820]  And then at the last minute we get Barton Gellman, not at the last minute, but we got Barton Gellman in as well.
[00:18.140 --> 00:24.820]  So now we have up Harlo Holmes, who is a Knight Mozilla Fellow at New York Times.
[00:25.400 --> 00:28.160]  Aurelia Moser, who is a Knight Mozilla Fellow at Ushahidi.
[00:28.340 --> 00:31.180]  Harlo also works for the Guardian Project, which does a lovely...
[00:31.180 --> 00:32.440]  Check out their secure tools.
[00:32.600 --> 00:34.700]  They're quite good for protecting you from things.
[00:34.940 --> 00:40.120]  And Barton Gellman, who has been one of the journalists working extensively on the Snowden documents.
[00:40.600 --> 00:41.360]  Take it away.
[00:43.020 --> 00:43.460]  Great.
[00:44.680 --> 00:45.380]  Thank you.
[00:50.040 --> 00:50.480]  Hi.
[00:50.860 --> 00:51.740]  I guess I'll start.
[00:51.980 --> 00:56.920]  We're gonna go down the line and do the typical intros to ourselves and what we do and why it's significant.
[00:56.980 --> 01:02.000]  for this talk, which is mostly going to be about dead drops.
[01:03.940 --> 01:04.890]  Dun, dun, dun.
[01:05.820 --> 01:07.290]  So my name is Aurelia Moser.
[01:07.540 --> 01:10.010]  I work with Ushahidi in Internews Kenya.
[01:10.170 --> 01:20.670]  So I work mostly out of East Africa and Ushahidi builds a suite of technologies that collect crowd-sourced information in places where penetration of smartphones isn't as ubiquitous.
[01:20.900 --> 01:22.880]  So we deal with a lot of feature phone data.
[01:23.020 --> 01:30.320]  We plot it to maps and allow journalists and technologists to read that information which has been encrypted or at least SSL secured.
[01:31.010 --> 01:37.880]  So that you can read what's happening in a space and a place and then write articles that are authoritatively informed by that information.
[01:38.260 --> 01:42.620]  So we build another suite of other technologies associated with that kind of crowd-sourcing platform.
[01:43.220 --> 01:49.080]  And then we also work a little bit with Internews Kenya, which has appendages all over the world.
[01:49.240 --> 01:58.880]  And they train data journalists in how to build technology and use technology both securely and then also just to do data reporting in their context of operation.
[01:59.300 --> 02:06.430]  So I work with them and somewhat with Harlo a bit as Open News Fellows from Mozilla.
[02:06.430 --> 02:09.710]  So we're going to go down the line and also do intros to ourselves.
[02:11.300 --> 02:11.980]  I'm Bart Gellman.
[02:12.320 --> 02:17.340]  I spent a long time as a national security investigative reporter at the Washington Post.
[02:17.660 --> 02:21.140]  I left in 2010 to pursue book projects.
[02:21.660 --> 02:33.840]  And when Edward Snowden came to me, initially through Laura Poitras, I decided I did not want to be freelance Bart working on that story.
[02:34.080 --> 02:35.300]  So I went back to the post.
[02:36.060 --> 02:43.160]  I've been, I'd say since 9-11, I'm not a coder, but I'm an early adopter.
[02:43.280 --> 02:58.060]  And I've been an enthusiastic customer of people like my fellow panelists here in terms of figuring out ways of securing my confidential notes, my plans, my sources, knowing that we're leaving more and more of a digital trail.
[02:58.700 --> 03:07.120]  If I had not known how to use GPG and Tor and encrypted Jabber, then I would not have been able to talk to Snowden.
[03:07.300 --> 03:09.580]  He was able to talk only that way.
[03:09.700 --> 03:14.240]  He was trying to drop a dime on the surveillance state without being surveilled, which is tricky.
[03:15.440 --> 03:31.440]  I have had to step up my game a little bit this past year because I, I've just, I've been, I've had more to protect and, and more people interested in peering over my shoulder than before.
[03:32.620 --> 03:45.060]  And I'm struck by the fact that even though I put a lot of time and effort into this, and I'm, this may be debatable, but I think I'm not a complete idiot about how to use this stuff.
[03:46.000 --> 04:02.720]  It is really way too hard to use, and there are way too many holes in the security, and there are, there are way too many sort of operational workflow issues that get overlooked when people are talking about the cool gadgets and the cool tools.
[04:05.240 --> 04:06.500]  I'm Harlo Holmes.
[04:07.240 --> 04:12.540]  I am a software developer, primarily with a group called The Guardian Project.
[04:13.020 --> 04:22.420]  We are featured in slide seven of one of the first NSA slide decks that got, that got leaked.
[04:22.600 --> 04:23.420]  And we were like, yeah!
[04:25.850 --> 04:26.420]  For...
[04:26.420 --> 04:28.800]  Which did not say they were pwned, which is also good.
[04:29.210 --> 04:43.090]  For, well, of the several applications that we develop at The Guardian Project, we are also one of the first developers on Orbot, which is now, you know, taken by the Tor Project, maintained by the Tor Project.
[04:43.850 --> 04:47.560]  But, it's been a pretty wild ride from that.
[04:48.600 --> 04:59.800]  We also have a suite of tools for secure communications, encrypted voice over IP, in the form of an app called OSTel.
[04:59.800 --> 05:03.500]  But this is actually only one of several offerings.
[05:03.800 --> 05:13.650]  Of course, maybe you guys are all familiar with Red Phone, which now is being, you know, beta tested, I guess, on iOS, where it was always on Android before.
[05:15.830 --> 05:28.840]  We have, we also have an app called Chat Secure, which is, you know, a chat app that does, you know, off the record, OTR, Jabber/XMPP communications.
[05:29.790 --> 05:37.540]  And we also have a suite of tools for multimedia, so like images and video.
[05:37.540 --> 05:52.220]  We have, most notably, a project called ObscuraCam, which came out a couple of years ago, which is like a very, very simple tool for just EXIF wiping and actual like pixelization or pixelation redaction within photos.
[05:53.130 --> 05:57.860]  And a really, really cool app geared towards citizen journalists.
[05:58.180 --> 06:01.520]  Well, not only citizen journalists, journalists in general.
[06:02.250 --> 06:04.400]  Also citizens, as it turns out.
[06:04.400 --> 06:11.980]  We are also citizens, called StoryMaker, which allows you to craft news stories, you know, on your mobile device.
[06:12.100 --> 06:13.960]  It has like an editor in the device.
[06:14.580 --> 06:27.150]  And it also includes lessons about how to create features for, you know, for publication in news organizations that are actually like make sense in terms of telling the story.
[06:27.300 --> 06:32.280]  And we wrap all of that in, you know, with, that all comes bundled with encrypted storage and stuff like that.
[06:32.280 --> 06:42.390]  Guardian Project also has a variety of developer libraries that we really, really, like, every time we go out, we encourage people to use and contribute to, submit patches, incorporate into your own projects.
[06:42.710 --> 06:51.800]  Most notably, as it was discussed here, IO Cypher, sorry, SQL Cypher, which is an encrypted port of, you know, the SQL libraries.
[06:51.800 --> 07:04.540]  Also IO Cypher, which is a package that you can use that provides the same amount of encryption, with a wrapper that looks exactly like Java.io suite.
[07:04.540 --> 07:06.710]  So for all of the developers there who want to use that.
[07:07.420 --> 07:13.800]  And a couple of other things that we tool on, you know, to make PGP available on Android devices, stuff like that.
[07:13.800 --> 07:15.400]  So please do check us out.
[07:15.650 --> 07:23.330]  Also, I'm working at the New York Times this year as a kind of in-house security dev with the Open News Foundation.
[07:23.580 --> 07:24.620]  Also with Aurelia.
[07:25.650 --> 07:26.340]  Yeah, totally.
[07:26.650 --> 07:27.980]  So before we get started...
[07:27.980 --> 07:28.120]  Oh.
[07:29.040 --> 07:29.720]  Oh, yeah.
[07:29.860 --> 07:33.840]  We just want to direct your attention to our lovely slides, which will not change.
[07:34.840 --> 07:39.340]  You can go to that URL, which has all of the links to tools and things.
[07:39.340 --> 07:44.020]  One thing that I always notice at talks is that you end up taking notes, or you don't take notes, and then you forget everything.
[07:44.540 --> 07:50.980]  And there's no reference to the people who are speaking if you come in late and you can't angrily, like, hate-tweet them when you disagree with what they say.
[07:51.560 --> 07:53.890]  So you can do that, as it turns out.
[07:53.980 --> 07:56.600]  So we'll keep this up, and we'll also talk a bunch of tools.
[07:56.770 --> 07:58.340]  There's some other things that couldn't be submoduled.
[07:58.360 --> 07:59.040]  They're down here.
[07:59.480 --> 08:01.210]  And then just general journal resources.
[08:01.540 --> 08:05.960]  We write a lot of guides, at least Internews does, for, say, for journal OPSEC practices.
[08:06.210 --> 08:07.680]  So you can totally check those out.
[08:07.680 --> 08:12.180]  There are also some other guides that we recommend or will probably reference at some point.
[08:12.480 --> 08:15.160]  So you can page through that as we're talking.
[08:15.540 --> 08:18.500]  We're going to kind of structure this pretty loosely as a conversation.
[08:18.560 --> 08:23.420]  But what we wanted to start out with is a few anecdotes of why operational security...
[08:23.420 --> 08:27.150]  I think we're all probably aware of why operational security is important for journalists.
[08:27.500 --> 08:39.500]  But there are several examples of not just, like, I guess I would say irresponsible OPSEC on the part of journalists, but also just information breach that they couldn't have protected themselves against.
[08:39.840 --> 08:51.620]  So we're going to talk through a few of those, just general examples, and then get into some scenarios where we would want to encrypt things, or would want to protect our information, and would want to securely drop information to our colleagues, but don't know exactly how to do that.
[08:53.520 --> 08:58.600]  So we're going to start off... I don't know if you guys want to start, if you want me to just...
[08:58.600 --> 09:01.660]  Oh, really quickly, a note about this GitHub repo.
[09:02.840 --> 09:10.820]  For better, for worse, actually, a lot of the developers whose projects we've included as sub-modules here are in this conference.
[09:10.820 --> 09:13.080]  So reach out, make friends, talk about code.
[09:13.280 --> 09:20.220]  Also, if we've included anyone here that for some reason does not want it to be shared in this particular way, please do let us know as well.
[09:20.760 --> 09:23.820]  Although, you know, commit history is commit history, so...
[09:23.820 --> 09:29.500]  Yeah, our superstar friends, really, and a lot of awesome people who contribute to really awesome projects.
[09:30.500 --> 09:36.720]  And also, that also goes for, you know, submit an issue if you find that there's some flaw with something we've recommended, or we've at least mentioned.
[09:36.920 --> 09:40.660]  We'll probably talk about those loosely, but we might not mention it on the GitHub repo.
[09:40.660 --> 09:44.660]  How that info is on that, is in that URL?
[09:45.200 --> 09:46.440]  At that URL, yes.
[09:46.440 --> 09:47.800]  Yeah, exactly, go check it out.
[09:47.900 --> 09:48.860]  At GitHub.com slash Mozilla, drop a like.
[09:48.900 --> 09:51.680]  Yeah, so you can keep it on your mobile or on whatever device.
[09:51.880 --> 09:54.000]  Exactly, yeah, that's why we gave it to you as a link.
[09:55.020 --> 09:55.380]  Cool.
[09:55.720 --> 09:58.460]  So, just traditional OPSEC issues.
[09:58.620 --> 10:11.320]  I mean, I think we all know maybe the Geraldo Rivera story of him being embedded in Iraq, like drawing the map of where he was embedded on video so that everyone could know where they were, where he was operating.
[10:11.580 --> 10:16.930]  But we also know of other examples of when, like, the John McAfee Vice Magazine photo...
[10:17.660 --> 10:18.720]  My favorite story ever.
[10:18.860 --> 10:21.100]  With the Geo metadata embedded in it.
[10:21.640 --> 10:33.980]  And then James Rosen phone call metadata was accessed, building card swipe metadata, the Chelsea Manning case, you know, compromised logging in ADM, right?
[10:34.060 --> 10:35.200]  ADM is a chat application.
[10:35.340 --> 10:41.240]  So even when you're using, you know, ostensibly secure technologies, there's still vulnerabilities that are pretty obvious.
[10:41.380 --> 10:44.680]  I don't know if you guys want to chime in with other examples of...
[10:44.680 --> 11:02.760]  Yeah, actually, what's interesting about operational security is, in addition to all of this tech that we're throwing up here, it's so inextricably linked with the legal system in the places that you're operating in, So, for instance, in the Rosen case,
[11:03.060 --> 11:06.360]  it actually is that when...
[11:06.360 --> 11:15.280]  Or in New York State, when a journalist would protect a source, like, you don't actually have to, like, hand over anything that you've collected.
[11:15.280 --> 11:21.000]  But if you move your information outside of the state, then you're, like, privy to other laws.
[11:21.160 --> 11:27.100]  And, of course, when you think about the cloud and, like, how documents are normally stored, like, that mucks up things even more.
[11:28.260 --> 11:28.600]  So...
[11:29.020 --> 11:31.420]  Bart, do you want to say anything, or do you want us to...
[11:32.780 --> 11:33.120]  Well...
[11:33.760 --> 11:39.220]  I guess one thing I wanted to say is, there's a lot of commentary post-Snowden about...
[11:39.580 --> 11:50.980]  And some of it's from Snowden himself, that investigative reporters are being reckless, irresponsible, clueless, if they are communicating only in the clear, if they're not learning to use these tools.
[11:51.360 --> 12:00.320]  And there are lots of people, a lot of different people now, with advice for journalists about how to use secure communication tools.
[12:00.360 --> 12:15.220]  And more of them need to do what Aurelia and Harlo are doing, which is to embed themselves in a newsroom, or spend some time in there to see how reporters actually work and have to work, and what their actual needs are.
[12:15.220 --> 12:20.800]  Because we're not going to be all sort of sitting around, you know, communicating only on pond.
[12:20.800 --> 12:22.380]  And, you know...
[12:23.790 --> 12:24.900]  We are...
[12:24.900 --> 12:26.080]  We are...
[12:26.080 --> 12:27.940]  We need to be open.
[12:28.080 --> 12:28.960]  We can't be off the grid.
[12:29.080 --> 12:30.420]  We're completely on the grid.
[12:30.560 --> 12:34.760]  We need to be open to communication by any channel, from anyone.
[12:35.060 --> 12:38.040]  We need to speak among ourselves.
[12:38.280 --> 12:39.640]  Newsrooms are open for a reason.
[12:40.460 --> 12:50.400]  There's a lot of sort of symbiotic and just sort of opportunistic information that crosses among us.
[12:50.400 --> 12:52.060]  And we get stories that way.
[12:52.280 --> 12:52.980]  So it's...
[12:52.980 --> 12:54.820]  On the one hand, we've got to be easy to contact.
[12:55.160 --> 12:56.420]  And it's got to be low threat.
[12:57.040 --> 13:07.660]  Because lots of people who you meet as a reporter don't start off thinking that they're going to be spooky sources of yours.
[13:07.980 --> 13:13.520]  So here's a scenario where I think technology can't help us an awful lot.
[13:13.720 --> 13:15.660]  But I'm looking for ways...
[13:15.660 --> 13:19.400]  And, you know, I hope any of you have any interesting ideas on this.
[13:19.400 --> 13:20.820]  So I go to a conference like this.
[13:21.500 --> 13:28.340]  And by accident, I'm seated next to someone who works for some entity that I want to know more about.
[13:28.560 --> 13:29.760]  And that's very secretive.
[13:29.860 --> 13:35.900]  I mean, essentially, my work as a reporter is to find out stuff that organizations and powerful people don't want me to know.
[13:36.040 --> 13:37.280]  That's the definition of the job.
[13:38.060 --> 13:40.280]  So I meet employee X.
[13:41.180 --> 13:43.300]  And we have a cup of coffee even.
[13:43.640 --> 13:45.580]  And we want to exchange contacts.
[13:46.200 --> 13:51.440]  And if I pull them aside and say, all right, from now on, you'll call me Popeye.
[13:51.440 --> 13:58.180]  And here's where you download Tails and we'll set up secret spooky accounts and encryption.
[13:58.520 --> 14:05.400]  It's as if I were saying, you know, here, let me have your phone number.
[14:05.620 --> 14:11.440]  And by the way, can you show me any recent STD tests and which brand of condom do you like?
[14:11.440 --> 14:14.340]  I mean, it's sort of, who are you?
[14:14.420 --> 14:15.000]  What are you talking about?
[14:15.100 --> 14:16.560]  I mean, I didn't agree to anything like this.
[14:16.740 --> 14:24.960]  And a lot of times when I'm talking to people who I regard as sources, they will even explicitly say to me, I'm not a source, right?
[14:25.020 --> 14:25.880]  I'm not breaking any rules.
[14:26.000 --> 14:27.260]  I'm not going to break any rules with you.
[14:27.740 --> 14:32.680]  And they tell me that over and over then while they gradually start moving towards the line.
[14:32.680 --> 14:38.740]  And as the line sort of starts to shift and they talk themselves into thinking, well, I didn't really tell him anything.
[14:38.740 --> 14:39.940]  He already seemed to know that.
[14:40.080 --> 14:41.100]  I'm just giving some context.
[14:43.260 --> 14:49.580]  Most of the most sensitive stuff I've gotten over a career as an investigative reporter is not from people who see themselves as whistleblowers.
[14:50.020 --> 15:07.240]  So how do we get to the point where I can offer any sort of meaningful protection using technology or communication security procedures where I can protect someone for whom there is already a trail showing that we've been in contact in more innocuous ways?
[15:07.240 --> 15:09.800]  That's one of the big problems I'm trying to solve.
[15:11.920 --> 15:12.520]  Okay.
[15:13.380 --> 15:16.920]  I actually... I'd like to...
[15:16.920 --> 15:18.660]  Oh, my gosh.
[15:18.800 --> 15:19.020]  What?
[15:19.620 --> 15:19.940]  Go ahead.
[15:19.940 --> 15:23.820]  I wanted to launch into like a discussion of some of the technologies.
[15:24.340 --> 15:31.260]  Like perhaps if we start with email as being like, yeah, the communication aspect.
[15:31.880 --> 15:32.840]  What is...
[15:32.840 --> 15:33.860]  What do you...
[15:33.860 --> 15:34.740]  What do you...
[15:35.600 --> 15:38.560]  What do you use in terms of securing email?
[15:38.860 --> 15:43.000]  And do you have any particular tips and tricks that you recommend?
[15:44.580 --> 15:45.060]  Yeah.
[15:45.240 --> 15:51.520]  Well, I mean, there is another use case, right, where someone does know from the beginning that they would like to talk to me securely.
[15:51.520 --> 15:54.360]  Someone who is very conscious of the risk.
[15:54.760 --> 16:05.480]  And the threat model is a capable adversary and not just the public relations spokeswoman who, you know, who wants to be in the room every time we talk or something like that.
[16:05.480 --> 16:17.700]  So in that case, I... for that kind of case, I published my GPG key as part of my Twitter profile and a Jabber address that I can use with OTR.
[16:17.860 --> 16:20.740]  So people who know how to use those tools can be in touch that way.
[16:21.500 --> 16:30.820]  I'm going to soon have a secure drop instance that comes to me, which makes it somewhat easier and makes it harder to screw up.
[16:32.440 --> 16:38.120]  Because so many of these tools work really well if you use them exactly right every single time.
[16:38.780 --> 16:45.180]  So suppose you create an email address for Popeye using Tor.
[16:45.840 --> 16:47.000]  So you're anonymized.
[16:47.260 --> 16:50.660]  And a hundred times in a row you log into it with a Tor browser.
[16:50.900 --> 16:52.060]  And then one day you're a little sleepy.
[16:52.200 --> 16:53.040]  You didn't have your coffee yet.
[16:53.100 --> 16:56.220]  And you go on your regular browser and you log into Popeye.
[16:56.320 --> 16:58.040]  And all of a sudden it's as if you never used Tor.
[16:58.040 --> 17:01.240]  It's as if you never were anonymous, that you are now tied to it.
[17:01.780 --> 17:10.360]  One of the great things that SecureDrop is doing, and I'm grateful for it, and that Tails does, makes it harder to screw up.
[17:10.460 --> 17:14.000]  It makes it harder to make those just sort of thoughtless errors.
[17:14.220 --> 17:15.480]  So, I mean, I publish keys.
[17:16.240 --> 17:18.280]  They're therefore associated with me.
[17:18.640 --> 17:22.920]  So at best one side of the conversation can be anonymous, can be private.
[17:22.920 --> 17:24.500]  But the other side is known to be me.
[17:25.280 --> 17:29.920]  And then using encrypted email, we might say, let's create accounts over here.
[17:29.940 --> 17:31.180]  And we'll start talking there.
[17:31.340 --> 17:33.040]  And now, neither side is known.
[17:33.420 --> 17:36.740]  That seems to be a fairly effective way to ensure privacy.
[17:36.960 --> 17:38.260]  You seem to...
[17:38.260 --> 17:41.080]  What's interesting about your comments there is that...
[17:41.840 --> 17:42.320]  Sorry.
[17:42.320 --> 17:44.820]  You seem to be encouraging, like, mindfulness.
[17:45.980 --> 17:50.380]  Like, the practice of, like, yeah, underscoring the fact that security is a practice.
[17:50.900 --> 17:54.560]  The very, very fact that, you know, these are steps...
[17:55.300 --> 18:11.700]  By using, you know, the Tor browser each time, you know, logging into your email, by, you know, having to kind of type out your PGP password to decrypt your email and things like that, these are steps that get put in your way.
[18:11.700 --> 18:14.020]  But every time you take...
[18:14.020 --> 18:21.880]  Or all of the time that you take in doing those steps actually, like, increases your mindfulness about the technologies and the security that you're giving yourself while doing it.
[18:22.320 --> 18:22.680]  Right.
[18:22.800 --> 18:24.560]  But it's also a giant tax on my time.
[18:24.700 --> 18:25.060]  Oh, yeah.
[18:25.200 --> 18:25.640]  It's all expensive.
[18:25.640 --> 18:34.320]  There are lots of things I'm doing now that make me less efficient as a reporter and writer.
[18:34.320 --> 18:36.940]  I'm making it harder for myself to get at my notes.
[18:37.100 --> 18:39.280]  I'm making it harder to get at my email.
[18:39.520 --> 18:50.380]  I'm making it so that if I'm on my smartphone, I can't read the email that people are sending me because I'm not willing to put my private keys on an Android or an iOS device.
[18:51.360 --> 18:52.160]  Things like that.
[18:52.160 --> 18:57.560]  And, you know, when I'm dealing with something as sensitive as Snowden documents, they're on an air gap machine.
[18:58.480 --> 19:05.500]  And the machine has been disabled in several different ways that make it a lot harder to get material from, you know, from place X to place Y.
[19:06.040 --> 19:07.720]  That's a feature, not a bug.
[19:07.900 --> 19:09.940]  But it takes a lot of my time.
[19:11.620 --> 19:20.520]  And I think something, if I can chime in briefly, a lot of what our educational materials on privacy and security focus on for internews, because we teach journalists in East Africa.
[19:20.640 --> 19:23.500]  We teach journalists all over the world how to secure their communications.
[19:23.740 --> 19:25.500]  And often we don't even teach tools.
[19:25.680 --> 19:32.460]  We teach really mindful operational security practices because those are easy to implement or not easy to implement.
[19:32.600 --> 19:33.420]  They're actually quite tedious.
[19:33.660 --> 19:36.960]  But implementing them everywhere doesn't require a certain infrastructure.
[19:36.960 --> 19:40.400]  It's not dependent on your accessibility to certain devices or things.
[19:41.660 --> 19:45.220]  And often those behavioral modifications are actually the most difficult.
[19:45.540 --> 19:48.320]  So people make fun of having code names for everything.
[19:48.660 --> 19:53.760]  But if you actually... the only reason that's worth making fun of is because it's so poorly implemented often.
[19:54.020 --> 19:56.240]  You know, if you implement it right, it actually is really effective.
[19:56.680 --> 19:57.960]  And people can't read your communication.
[19:58.120 --> 20:02.400]  It doesn't involve steganography or any kind of, you know, stylometric scrambling.
[20:02.660 --> 20:06.560]  It just involves being mindful of what you're doing with your data and your information.
[20:07.400 --> 20:10.560]  So a lot of our guides focus more on those behavioral practices.
[20:10.860 --> 20:12.660]  And also just not logging into social media.
[20:12.960 --> 20:15.200]  And not, you know, being... not being proud about your location.
[20:15.400 --> 20:18.720]  Or about broadcasting the fact that you as a journalist were tied to a particular communication.
[20:19.400 --> 20:20.820]  Or tied to a story even.
[20:20.960 --> 20:23.260]  Being like an anonymous journalist in Damascus, for example.
[20:23.440 --> 20:24.900]  And signing all of your articles that way.
[20:24.940 --> 20:27.780]  Instead of saying, I have to be this person with a persona.
[20:28.160 --> 20:29.540]  Because a persona is dangerous.
[20:31.380 --> 20:37.200]  At the times, actually, when I had gone up to, you know, the guys in security.
[20:37.400 --> 20:39.940]  And asked them about, like, you know, what exactly...
[20:39.940 --> 20:41.300]  Like, what's in your handbook?
[20:41.480 --> 20:42.380]  What's in your rule book?
[20:42.520 --> 20:44.260]  You know, day one, when a journalist comes in.
[20:44.500 --> 20:45.420]  You, you know, sit him down.
[20:45.500 --> 20:47.280]  You say, like, alright, you're gonna install this software.
[20:47.400 --> 20:48.920]  And you're gonna, like, establish this practice.
[20:49.270 --> 20:51.080]  And such a thing actually does not exist.
[20:51.080 --> 20:56.260]  I mean, and I, I, I don't want to think that I'm reading into the answer a little bit.
[20:56.660 --> 21:01.700]  But, I mean, ultimately, for an organization as large as the times.
[21:01.740 --> 21:05.940]  Which has, you know, quite a reputation to withhold and to protect.
[21:07.040 --> 21:13.740]  You can't actually advocate for one particular tool officially.
[21:14.240 --> 21:17.240]  Simply because, you know, there's so much liability at stake.
[21:17.240 --> 21:18.600]  If any of that goes wrong.
[21:19.860 --> 21:23.080]  Also, like, I, once again, it is about mindfulness.
[21:23.380 --> 21:24.280]  It's about practice.
[21:24.560 --> 21:28.580]  It's about giving people, like, teaching tools.
[21:29.080 --> 21:31.020]  And giving them options for tools.
[21:31.400 --> 21:34.840]  But, ultimately, giving them the confidence to use these tools.
[21:35.080 --> 21:38.300]  Because you, your situation, or your threat model, or whatever.
[21:38.520 --> 21:41.420]  That changes from scenario to scenario, day to day.
[21:41.620 --> 21:43.460]  Depending on who you're talking to and where you are.
[21:43.820 --> 21:48.500]  And so, yeah, once again, people need the confidence in order to make choices about their security.
[21:48.500 --> 21:49.100]  On the fly.
[21:49.560 --> 21:50.780]  And that's the most important part.
[21:53.220 --> 21:53.580]  Right.
[21:53.760 --> 21:57.200]  And I think with that, too, we'll move into some more specific scenarios.
[21:57.580 --> 21:58.760]  I think we've talked loosely.
[21:58.960 --> 22:00.580]  You've already mentioned a few from your work.
[22:01.240 --> 22:05.660]  But we were going to talk through some examples of, you know, when you will want to secure your communications.
[22:05.660 --> 22:09.020]  Or your files and store them appropriately and securely.
[22:09.680 --> 22:10.440]  What do you do?
[22:10.720 --> 22:14.480]  And most of this, too, the reason we set this up kind of as a versionable repository.
[22:14.820 --> 22:16.380]  And also just linked to a bunch of other tools.
[22:16.520 --> 22:18.280]  Is that we are hoping for feedback from people.
[22:18.420 --> 22:19.500]  Not just rage tweets.
[22:20.320 --> 22:22.140]  But also just people...
[22:22.140 --> 22:23.000]  You get a lot of these rage tweets?
[22:23.000 --> 22:23.560]  No.
[22:23.780 --> 22:24.000]  No, no, no.
[22:24.000 --> 22:24.820]  This isn't all personal.
[22:25.160 --> 22:26.220]  Nothing is personal for me.
[22:27.220 --> 22:27.580]  Yes.
[22:28.020 --> 22:28.320]  So...
[22:29.800 --> 22:36.120]  But if you want to submit an issue or, like, submit another tool that you know about that you think would be useful, we have it somewhere in the repository.
[22:36.120 --> 22:38.100]  I think there's a laundry list of what do you do for browsing?
[22:38.200 --> 22:40.480]  What do you do for secure, you know, operating systems?
[22:40.660 --> 22:41.920]  What do you do for email?
[22:42.260 --> 22:43.120]  You know, what would we recommend?
[22:43.280 --> 22:44.080]  What are the tools we use?
[22:44.180 --> 22:48.320]  What is a suite of chat applications that are available that supposedly secure your communication?
[22:49.260 --> 22:51.180]  And there are all kinds of other rankings.
[22:51.420 --> 22:59.920]  There's a few other journalism resources that we link to that provide those kind of rankings of up-to-date, you know, has this been audited yet?
[23:00.200 --> 23:02.960]  Here's the link to the website that tells you if it's been audited yet.
[23:04.220 --> 23:05.020]  So, yeah.
[23:05.140 --> 23:09.320]  I wanted to direct everyone to that and say please contribute to it because we're not perfect.
[23:09.540 --> 23:14.780]  I'm sure there are many things we could introduce to our users and our fellow journalists that would be incredibly useful.
[23:15.580 --> 23:21.400]  But, yeah, so, first scenario, we kind of talked about getting data out of a...
[23:21.400 --> 23:30.020]  If you're operating remotely, for example, and corresponding with a source or just operating as an investigative journalist embedded somewhere else.
[23:30.240 --> 23:34.940]  I work mostly with training journalists who are already in a particular location.
[23:34.940 --> 23:40.840]  So, they grew up in Eritrea and they want to be journalists despite the fact that it's illegal and they could die.
[23:41.320 --> 23:43.920]  They're interested in reporting about their own context.
[23:43.920 --> 23:52.000]  But there are also, you know, journalists who travel and do story or are embedded in a particular location and write stories about that location and what's going on.
[23:52.100 --> 23:58.880]  And how do you correspond with your sources safely and then return to your home country without having to worry about all of your things being compromised?
[23:59.460 --> 24:02.320]  That includes your notes and your hardware and everything.
[24:02.320 --> 24:07.860]  So, I think you had a more specific anecdotal example of that that maybe everyone's aware of, maybe.
[24:08.380 --> 24:10.600]  I went to visit Snowden in Moscow.
[24:10.880 --> 24:12.160]  That's what you came to hear.
[24:14.480 --> 24:15.440]  In December.
[24:18.480 --> 24:25.080]  So, one of the things was, how do I arrange to meet without being followed by everyone in Moscow?
[24:25.720 --> 24:36.140]  I mean, one of my threat models was other journalists who have spent a long time, all the Moscow-based journalists and a bunch of visitors, trying to find him, figure out where he is.
[24:36.140 --> 24:37.200]  He didn't want to be found.
[24:38.200 --> 24:50.640]  There were also people who, if they were aware of his presence, would have tipped off the Russian media and had camera crews waiting for me when I arrived, which I didn't really relish.
[24:52.100 --> 24:58.200]  And there are people who are, you know, sort of purport to be on his side but are not reliable.
[24:59.360 --> 25:02.060]  So, we had to work out a way to meet.
[25:02.060 --> 25:11.640]  And all I knew was one place I was going to be at a certain time and then I was going to get a signal about where the other place would be and so forth.
[25:11.640 --> 25:15.600]  But my bigger problem was when I was done.
[25:15.700 --> 25:20.580]  I now have, I spent 14 hours of interviews with him.
[25:20.740 --> 25:24.420]  I had photos, I had audio, I had a lot of text.
[25:25.800 --> 25:29.580]  And I did not want to hand that either to the U.S. government when I arrived there or to the Russian government when I departed.
[25:35.360 --> 25:37.500]  And so, simple problem.
[25:38.220 --> 25:39.920]  How do I keep this material?
[25:40.080 --> 25:50.660]  How do I maintain control of it so that I'll have it when I get back and not hand it to anybody else in a way that they could have access to?
[25:50.660 --> 25:56.260]  And I asked a lot of very competent security people this question, looking for advice.
[25:56.460 --> 26:02.400]  And it was surprising how hard it was to get an answer where someone said, well, clearly, here's the ten things you do.
[26:02.760 --> 26:06.440]  Just one through ten, that's your list, you know, you're done.
[26:07.620 --> 26:09.400]  That doesn't exist, as far as I know.
[26:10.000 --> 26:26.840]  And so, I did the second best thing, which actually is part of the skill set of a reporter, which is interview lots of experts and weave together an answer to sort of master the information and simplify it and figure it out.
[26:26.920 --> 26:29.980]  And so, I came up with a method, and I'll tell you some of it.
[26:34.000 --> 26:37.260]  I wanted to do it anonymously, so I worked over Tor.
[26:38.480 --> 26:43.920]  I encrypted to... I did not bring any private keys with me.
[26:44.040 --> 26:46.640]  I didn't bring any key material with me to Russia.
[26:46.660 --> 26:51.260]  I encrypted to a public key that I controlled, but I didn't want...
[26:51.260 --> 26:58.260]  but it's because the way GPT works, the person to whom it's encrypted is discernible to an adversary.
[26:58.520 --> 27:05.820]  So, I then put it inside a symmetrically encrypted package that did not reveal that information.
[27:06.380 --> 27:13.260]  And I stored it inside still another encrypted container using different material.
[27:13.260 --> 27:21.720]  And I kept some of it on teeny little micro SD cards and conveyed them back physically to where I was going.
[27:22.880 --> 27:30.500]  And some of it I sent to sort of multiple anonymous online temporary storage facilities.
[27:30.900 --> 27:33.000]  And I hoped that I would remember all the passphrases.
[27:34.240 --> 27:39.240]  Because I have locked myself out before, sometimes temporarily and sometimes not.
[27:39.300 --> 27:44.520]  It would kind of suck to get it all out securely and not be able to open it when I got back to New York and write my story.
[27:46.000 --> 27:58.300]  And it all worked, but it took a huge amount of time for me to sort of work out how I was going to do that and then create accounts in ways that would not be linked to me.
[27:58.820 --> 28:11.520]  Because I don't want that stuff to be pulled down by a foreign government or by any of you, for that matter, who say, yeah, look, I can obviously pwn a journalist.
[28:13.060 --> 28:16.380]  And I wanted to keep it and keep it secure.
[28:16.640 --> 28:19.400]  And I would love to have better ideas for how to do it next time.
[28:19.580 --> 28:27.280]  One thing you know you don't want to do based on experience that is now publicly known is you don't want to give it to your significant other to carry through Heathrow.
[28:29.080 --> 28:30.980]  You don't want to send it by FedEx.
[28:34.100 --> 28:37.100]  I mean, there's lots of good stories about how not to do it.
[28:37.400 --> 28:38.940]  There aren't a lot of good stories about how to do it.
[28:40.540 --> 28:42.360]  I think, too, that's one thing.
[28:42.680 --> 28:46.340]  Harlo and I have talked about this a little bit, and I think the Tao Center is doing some work on this.
[28:46.560 --> 28:53.780]  You might have gotten the... there was a book that was printed along with all the stickers for OpenITP and such that was out for digital security for journalists.
[28:54.060 --> 29:07.540]  And we're kind of talking about this general need in journalism programs for some formalized operational security education and presenting students with, you know, complex problems where you have multiple media types and you need to get it from one location to another.
[29:07.660 --> 29:14.020]  How do you model, you know, the various threats along that trajectory and figure out what are the most secure ways to store that information?
[29:14.020 --> 29:17.480]  If it's digital, then there are a whole suite of ways to do it that way.
[29:17.480 --> 29:22.960]  If it's not, if it's physical, you know, how do you transport physical materials or duplicate them or is that advisable?
[29:23.180 --> 29:25.040]  And nobody really is ever educated in that.
[29:25.140 --> 29:30.480]  You kind of ad hoc learn or you find some security friends and then they help you, but everyone has a different methodology.
[29:30.820 --> 29:38.720]  So kind of institutionalizing that in journalism education, I think, is a real problem and gap and something that needs to happen.
[29:39.460 --> 29:46.760]  But yeah, if you have suggestions for that scenario and want to submit them, I'm sure he can give you a reference and coding.
[29:46.760 --> 29:50.760]  Actually, I'll give you one very practical problem that I have not solved to my satisfaction.
[29:50.820 --> 29:59.260]  So I have... because there's a... happily, there's a blossoming of tool creation now.
[29:59.900 --> 30:11.480]  And because people use different things, I have identities on a bunch of different services, not all of which I trust as much as others, but they're what people use.
[30:11.480 --> 30:28.520]  So I have Silent Circle and I have Redfin and I have SureSpot and Threema and Wicker and, you know, and multiple Jabber accounts because if it's someone very sensitive, I don't use the same account for two different people.
[30:29.780 --> 30:32.580]  And it gets to be a very big number of things.
[30:32.720 --> 30:34.080]  And how do I remember all those passphrases?
[30:34.620 --> 30:41.020]  And I am very unsure what's a safe way to use...
[30:41.540 --> 30:44.860]  What password manager, if any, would be safe to use?
[30:45.280 --> 30:48.280]  Should I GPT encrypt a text file?
[30:48.280 --> 30:53.160]  Well, if I do that, then every time I open it, the whole thing is there in clear text on a computer.
[30:54.020 --> 30:58.440]  Or do I have 10,000 little teeny little text files, each of which is GPT encrypted?
[30:58.840 --> 31:09.300]  I mean, there's, you know, you could fall all the way down a rabbit hole in this stuff and I don't have a good answer that is reasonably usable and secure.
[31:12.430 --> 31:14.590]  Carla, do you have anything you want to say or should we move on?
[31:15.330 --> 31:15.930]  I don't know.
[31:16.410 --> 31:30.710]  Right, so we kind of already addressed the what do you do when you correspond with a new source and I don't know if you wanted to go into any further, like how you initiate that correspondence or what have been the various types of source correspondence initiation you've engaged in.
[31:31.530 --> 31:33.230]  But if not, we can move on to another one.
[31:33.450 --> 31:37.510]  Do you have any additional thoughts, either of you actually, in case you've corresponded with...
[31:38.810 --> 31:40.170]  Well, here's another basic problem.
[31:40.170 --> 31:41.230]  How do you arrange to meet?
[31:41.750 --> 31:44.390]  Face-to-face could be, by far, the best way to do something.
[31:46.830 --> 31:51.110]  But, you know, you're probably not going to travel somewhere and just show up at a door.
[31:51.270 --> 31:56.790]  Actually, sometimes I do, because that seems like the safest thing to do or because I think someone will say, no, go away.
[31:57.050 --> 32:01.150]  And it's a little bit harder to do that face-to-face than it is on the phone.
[32:01.790 --> 32:03.750]  But most of the time, you have to arrange to meet someone.
[32:03.890 --> 32:09.550]  And most of that is done by, you know, some form of telecommunications or another.
[32:09.550 --> 32:13.070]  What's a good way to do that without leaving a metadata trail?
[32:14.450 --> 32:18.110]  As Snowden, I think, said yesterday, metadata doesn't lie.
[32:19.690 --> 32:26.950]  You can know that you've got a code where you say on the phone, let's meet Thursday at the blue spot.
[32:27.110 --> 32:30.030]  And both of you know that means we're meeting Tuesday at the red spot.
[32:30.030 --> 32:34.510]  But the fact that I call the other person is really hard to conceal.
[32:35.570 --> 32:37.830]  So I've, you know, I've tried different things.
[32:37.970 --> 32:43.810]  I mean, sometimes I tell a mutually trusted connection to pass a message along.
[32:45.670 --> 32:51.810]  And preferably in person, because otherwise it's just one hop away and not really that much protection.
[32:52.710 --> 33:02.230]  But again, I'm always looking for suggestions on how to initiate, how you bootstrap from an insecure method of contact to a secure one.
[33:04.190 --> 33:19.950]  There are also, I mean, if you want to do things digitally, there are interesting ways to pass messages along that, you know, are just kind of fun and have to do with like, you know, the way that we use the web in and of itself.
[33:19.950 --> 33:39.890]  So for instance, I've actually seen, or I've heard anecdotes of just using the fact that like Facebook is so, or what people share on Facebook, using that as like, as a way to kind of obfuscate the fact that you're, you know, you've, you want someone to see a,
[33:40.050 --> 33:45.930]  or visit a certain website that might, or URL that might have, you know, information specific for that person.
[33:46.910 --> 33:59.870]  But instead of communicating that through a way that is pinpointable through metadata, meaning that like person A talked to person B, you know, the, if that information is like passed through Facebook through, you know, this network of, of interconnected people,
[34:00.110 --> 34:04.790]  you have a way of, of getting your original message to its intended recipient.
[34:06.310 --> 34:10.330]  Right, I mean, I could, I could find a way to use the word spinach in a tweet.
[34:10.790 --> 34:13.750]  That means Popeye is supposed to come, you know, find me, but I mean...
[34:15.410 --> 34:16.290]  Now we know.
[34:19.090 --> 34:19.530]  Spinach.
[34:19.790 --> 34:21.470]  I have to change all my codes now.
[34:23.810 --> 34:25.930]  Yeah, but they're, but they're awkward.
[34:26.110 --> 34:27.290]  And then you have to set them up.
[34:27.430 --> 34:28.850]  And the person has to be willing to set them up.
[34:29.010 --> 34:30.890]  And in my world, that's seldom the case.
[34:34.370 --> 34:35.310]  Yeah, great.
[34:35.790 --> 34:43.730]  So another, so maybe some more just general examples of reasonable source anonymity security.
[34:44.170 --> 34:45.890]  Like how do you maintain source anonymity?
[34:45.970 --> 34:48.370]  How do you maintain reporter anonymity?
[34:48.610 --> 34:57.990]  And, let's see, another one that we had was, oh, I guess we already sort of discussed that publication security.
[34:57.990 --> 35:12.190]  I mean, like, you know, if you're in the middle of something heavy and like, you don't want ISIS or Mossad or somebody like that to find out, oh, reporter over here.
[35:21.850 --> 35:22.450]  Yeah.
[35:23.790 --> 35:38.830]  I mean, of course, we all hear some pretty, pretty scary stories about the, just the very idea that having a communication device on you gives you a signal that is pinpointable and that makes you a target.
[35:38.830 --> 35:51.870]  And I actually, I'm not really comfortable like saying like, oh, well, you know, you should do this in order to like totally obfuscate the fact that you have like a communication device on your, because of all of the things that we discuss.
[35:52.370 --> 35:58.130]  There are always holes in them and someone always gets hurt other than like not leaving it.
[35:58.310 --> 36:00.130]  But then I like leave your device at home.
[36:00.350 --> 36:12.470]  But then again, if you are already under suspicion, the very idea or like just the very fact that you're not, that you're walking around, not connected, not carrying a device and not like leaving a pattern that is noticeable.
[36:12.470 --> 36:21.170]  That also like is totally reason for somebody to come and like knock you over the head because you're not acting as they expect you to act.
[36:21.270 --> 36:22.370]  You're not fulfilling that pattern.
[36:22.510 --> 36:24.430]  So it's very, very difficult to know these things.
[36:25.310 --> 36:34.630]  Yeah, I was going to say too, I guess I can't really speak to that in terms of embedded journalists except for myself who was very conspicuous in Nairobi, for example, where I was working before.
[36:35.450 --> 36:41.930]  But I mostly worked to train journalists who are less conspicuous who grew up in that area, you know, who maybe aren't out as journalists.
[36:43.290 --> 36:43.970]  Oh, great.
[36:44.190 --> 36:45.170]  So maybe we should wrap up.
[36:45.310 --> 36:46.770]  But I'll just finish this last thought.
[36:46.930 --> 36:53.230]  So I mostly deal with securing people who aren't obviously foreign in a space.
[36:53.230 --> 36:59.410]  But, and that security is a whole other host of other issues and not necessarily technologically related.
[36:59.610 --> 37:05.630]  You know, you're just training them in operational security, but even just publishing something openly, which is very journalistic.
[37:05.810 --> 37:08.850]  We talked about how we're, you know, our mandate is mostly open information.
[37:08.950 --> 37:14.090]  So converse to that is our desire to like make sure that we are safe ourselves.
[37:14.390 --> 37:21.370]  And a lot of the journalists that I work with compromise that every time they publish a story, even anonymously, you know, they live in a space.
[37:21.370 --> 37:26.470]  And they're actually the most vulnerable in places where security is the most fragile.
[37:27.010 --> 37:35.910]  Local journalists, not necessarily international transplants, because international transplants usually come with a host of other security, like an entourage of people protecting them.
[37:36.050 --> 37:37.790]  And then there would be outcry if they go missing.
[37:37.990 --> 37:43.490]  But in most places in the world where journalism isn't very welcome, people go missing all the time.
[37:43.490 --> 37:50.530]  the disappeared happened and people just pretend that it's not happening or they're told to pretend it's not happening to protect their own safety.
[37:51.550 --> 37:52.030]  Right.
[37:52.190 --> 37:54.290]  But that's a bit of a tangent more related to my work.
[37:54.410 --> 37:55.810]  So maybe Harlo wants to bring it back.
[37:56.130 --> 38:06.910]  There's actually a growing trend among smaller publishers to kind of train their journalists to use a dead man switch.
[38:06.910 --> 38:11.150]  So like, and this is actually really, really big in Mexico.
[38:11.410 --> 38:15.410]  There's an app that has been under development called Hansel.
[38:16.050 --> 38:20.710]  And it's been under development for like about three years really in like various stages.
[38:21.230 --> 38:23.610]  Now produced by Amnesty International.
[38:23.870 --> 38:29.610]  That is for the, you know, in response to the fact that like journalists get kidnapped all the time.
[38:29.610 --> 38:37.330]  And so being able to check in with your editor or with your desk and also to like kind of have a panic button if something goes wrong.
[38:37.570 --> 38:45.110]  So like just with the press of a button, you can just send out like spew out messages that contain your current location.
[38:45.430 --> 38:52.390]  You know, and that also includes like not just your latitude and longitude, but also like the cell towers that you're associated with.
[38:52.390 --> 38:54.070]  Because of course those can be looked up.
[38:54.490 --> 39:05.530]  And also even like Wi-Fi networks because like SSIDs or BSA, more likely BSSIDs are actually really, really great at corroborating location.
[39:06.010 --> 39:14.930]  So, and these are like tricks that we took from like the ad industry because like, you know, that's what they're using to track our, how we're walking through a shopping mall.
[39:15.350 --> 39:22.030]  But these are also like techniques that we, you know, that are used and bundled into, you know, potentially life-saving technologies.
[39:23.090 --> 39:29.450]  But if you, and I know you're not, you're not saying this, but I mean, I have talked to people who think, you know, well, all you have to do is this.
[39:29.690 --> 39:31.150]  And then you saw, you know, problem solved.
[39:31.370 --> 39:39.710]  And, you know, it's one thing to put up the bat signal and another thing, is Batman there waiting to come and save you?
[39:40.970 --> 39:45.270]  And, you know, the Washington Post, if it ever had one, lost its Batman.
[39:47.310 --> 39:51.290]  And likewise, there's, you know, I think it was one of the buyouts back in the 90s.
[39:55.250 --> 39:56.650]  So, there's always trade-offs.
[39:56.790 --> 40:01.350]  I mean, as a consumer, I desperately want someone to say, all you have to do is this, this, and this, and you're safe.
[40:02.210 --> 40:10.290]  Just in terms of the physical security that you were alluding to, sir, I spent a couple of months in Iraq during the last war.
[40:11.090 --> 40:16.190]  And it was late 03, early 04.
[40:16.450 --> 40:18.230]  I was hunting for weapons of mass destruction.
[40:18.450 --> 40:19.630]  And I didn't find any.
[40:21.690 --> 40:29.390]  And I, it was about the time it was starting to get dangerous for Westerners to move around there.
[40:29.390 --> 40:39.530]  And also, one day I was told that there was someone who was not me, who was claiming to be Bart Gellman and asking a lot of questions like an intelligence officer type would ask.
[40:39.730 --> 40:45.250]  And that made things more dangerous for me to show up as the second Bart Gellman in the neighborhood.
[40:45.530 --> 40:47.210]  The artist formerly known as...
[40:47.210 --> 40:49.170]  Can I interject one thing?
[40:49.350 --> 40:50.430]  And I'm sorry to do this.
[40:51.070 --> 40:54.710]  That's actually a really, really interesting thing, the type of squatting problem.
[40:54.970 --> 41:01.990]  Like, for instance, my name is spelled H-A-R-L-O, but there is a common spelling of my name, which is H-A-R-L-O-W.
[41:02.250 --> 41:11.330]  And actually, because I have, like, you know, Harlo.Holmes at New York Times, I, I've heard, like, you know, people saying, like, why didn't you respond to that email?
[41:11.630 --> 41:13.170]  Like, I never got your email.
[41:13.290 --> 41:15.690]  And of course, they're sending it to, like, you know, into the ether.
[41:15.690 --> 41:18.870]  But if, it's the same thing on Gmail as well.
[41:19.230 --> 41:30.510]  So, like, there are people out there who have, or there is someone out there who has Harlo with a W.Holmes at gmail.com, who actually gets, like, a lot of stuff that is intended from my eyes only.
[41:30.810 --> 41:35.230]  And yet, I have no recourse because Google does not provide that protection for people.
[41:37.150 --> 41:39.270]  All I was going to say, just in turn, and we'll go to questions.
[41:39.290 --> 41:40.250]  I have doppelgangers.
[41:40.450 --> 41:41.470]  My name is too weird.
[41:42.410 --> 41:46.010]  But just in terms of the trade-off thing, that, you know, do you carry your phone or not?
[41:46.110 --> 41:48.410]  If you carry your phone, it could be tracking your location.
[41:48.550 --> 41:51.670]  If you stick it in a Faraday bag, it looks really suspicious if you get stopped.
[41:51.870 --> 41:56.570]  If you don't carry the phone, then you can't communicate when you might really need to.
[41:57.410 --> 42:05.470]  The point of the Baghdad story was, right around that time, some reporters were trying to blend in and not be noticeable as reporters.
[42:05.470 --> 42:13.890]  Others were giving up and driving around in really expensive armored vehicles with, you know, no flat tires and that kind of thing.
[42:14.570 --> 42:18.710]  I prefer inconspicuous, so I got, like, the most beat-up, crappy local car.
[42:18.790 --> 42:24.290]  I could get it and let my beard grow out and just hope not to be conspicuous as a journalist from a distance.
[42:24.290 --> 42:25.190]  But there are big trade-offs.
[42:25.310 --> 42:29.590]  I had a friend who went out a month later and got waylaid in the road.
[42:29.590 --> 42:39.450]  And if he hadn't been in an armored vehicle, then the three or four Kalashnikov clips that got emptied into his car would have been, you know, a whole lot worse.
[42:39.630 --> 42:40.770]  Instead, he got to walk away from it.
[42:44.370 --> 42:45.090]  Oh, totally.
[42:45.090 --> 42:46.790]  We would love to take your questions, please.
[42:46.790 --> 42:47.890]  We would love to take your questions.
[42:48.010 --> 42:48.430]  Mike, sir.
[42:48.890 --> 42:48.990]  Yeah.
[42:50.110 --> 42:50.810]  There's a microphone.
[42:50.950 --> 42:58.810]  I was going to say, too, just as a side note, unfortunately, when I travel and do anything, I can't really grow out my beard and pass.
[42:59.370 --> 43:00.170]  And pass.
[43:00.830 --> 43:02.810]  So women are very conspicuous.
[43:02.910 --> 43:06.050]  And women investigative journalists, too, you take a lot of risks when you're traveling.
[43:06.630 --> 43:17.270]  And you put a lot of people in danger, like the entourage of people in my vicinity trying to protect me in any given context where I might not be expected as a woman doing any kind of work, for example.
[43:19.950 --> 43:23.670]  It's quite a human cost and a lot of time investment and cost.
[43:23.830 --> 43:41.370]  So sometimes it's almost... I often think that educating journalists in those contexts is even better, from my perspective, because then they can go out and pass and interview people inconspicuously, whereas I would require so much more time and money and investment to protect me from whatever might happen.
[43:41.370 --> 43:42.250]  Right.
[43:42.470 --> 43:43.390]  So that was just an aside.
[43:43.610 --> 43:44.290]  But first question.
[43:44.530 --> 43:45.870]  Thanks for a wonderful talk.
[43:46.650 --> 43:50.170]  So there's been a big debate recently that I personally disagree with.
[43:50.330 --> 43:52.790]  But the debate is, what is a journalist in the U.S.?
[43:52.790 --> 43:58.290]  When the Snowden story broke out, a lot of commentators said that Glenn Greenwald wasn't a journalist.
[43:58.310 --> 43:59.090]  He was a blogger.
[43:59.330 --> 44:04.410]  Alexa O'Brien, who covered the Manning trial single-handedly, was often called a blogger.
[44:04.410 --> 44:06.070]  The New York Times later corrected that.
[44:06.650 --> 44:11.910]  Personally, I'm a First Amendment attorney and I represent clients who are often called...
[44:11.910 --> 44:14.470]  The journalists who are called other things than journalists.
[44:14.730 --> 44:29.470]  And my question is, this talk largely focused on journalists, but does that necessarily play into this debate and create a barrier that shouldn't be there when journalism is really a practice that's engaged in by everyone?
[44:30.210 --> 44:31.690]  I completely agree with you.
[44:31.710 --> 44:33.410]  I think journalism is a function.
[44:33.750 --> 44:40.110]  I mean, if you are gathering information for publication for other people to share, that's it.
[44:40.250 --> 44:41.950]  I see no distinction.
[44:42.170 --> 44:44.990]  I wouldn't claim any special right or privilege.
[44:45.070 --> 44:53.390]  The idea that we would license journalists either explicitly or by pretending that we can decide who's what seems un-American to me.
[44:54.230 --> 44:56.510]  Also, I mean, that's just the American...
[44:56.510 --> 44:57.130]  Yes, please.
[44:57.950 --> 44:59.130]  Give us a big applause.
[45:02.010 --> 45:10.330]  There are other countries... I mean, of course, this is... you're speaking to the American context, but there are other countries where, like, actually, it's clearly defined that, like, if you have...
[45:10.330 --> 45:18.870]  If you have a blog and it gets more than 300 page views, then you're actually... congratulations, you're a journalist, and now you're going to jail.
[45:21.590 --> 45:22.310]  So, yeah.
[45:22.730 --> 45:23.910]  It's a good news, bad news kind of situation.
[45:23.970 --> 45:24.450]  Yeah, exactly.
[45:24.630 --> 45:26.170]  So, the good news is you're a journalist.
[45:26.170 --> 45:27.270]  The bad news is...
[45:28.670 --> 45:37.470]  And also, the way that journalism inside the United States is changing, just as an industry, is going to be very, very interesting in how we define these roles.
[45:37.690 --> 45:44.390]  So, for instance, like, I don't know, like, if you're, like, writing for Gawker or whatever, like, you are a journalist, but, like, are you a journalist?
[45:44.490 --> 45:45.150]  Are you a blogger?
[45:45.650 --> 45:51.370]  You can, of course, like, as a Gawker user, like, set up your... you know, everyone has, like, the Kinja account, right?
[45:51.610 --> 45:56.390]  And so, anything that you put... and they want you to treat Kinja like live journal, pretty much.
[45:56.590 --> 46:05.670]  So, if you're live journaling on Gawker's website, there is a chance that, like, if that is... if that gets enough views, that's going to be promoted to one of their sites.
[46:05.830 --> 46:10.230]  So, like, Gawker, or Jezebel, or, you know, like, Gizmodo, and that makes you a journalist.
[46:11.190 --> 46:16.690]  And, I mean, we haven't yet seen when, like, defamation suits come down, but, like, I mean, that is definitely a possibility.
[46:17.670 --> 46:38.930]  So, like, in terms of these technologies, it actually... it behooves us that, like, you know, we have, like, this right here, and a lot of other things that are not on this list or whatever, because these tools are, you know, kind of lightweight enough for you to take control of your own security if you're putting yourself in that situation.
[46:39.730 --> 46:43.710]  Yeah, and, too, like, at Ushahidi, at least, we build a lot of crowdsourced technologies.
[46:43.910 --> 46:53.990]  They're all open-source, and they're meant for people to become citizen journalists and kind of take charge of their own deployments and deploy them wherever they're located and collect citizen information from all over.
[46:55.870 --> 47:08.290]  So, yeah, so we have our crowd map and open-source mapping technologies, but we also have, like, Ping, which came out of Westgate and was meant to be able to ping your network and find out where your family is if something happens, like a mall attack,
[47:08.670 --> 47:10.330]  a horrific mall attack.
[47:10.530 --> 47:18.490]  And we have CrisisNet, too, which aggregates a bunch of social media information about crisis all over the world, and you can ping it from one unified API, which is nice.
[47:18.810 --> 47:27.490]  But we have a bunch of open-source technologies that's meant to enable people who are, you know, bloggers, who are journalists, actually, because they're the most valuable if they're on the ground.
[47:28.950 --> 47:38.010]  Yeah, I just wanted to ask, earlier you mentioned that you look suspicious to a strong adversary if you use these OPSEC tools.
[47:38.250 --> 47:42.450]  So, if you do everything in the clear, they can see everything you're doing.
[47:42.550 --> 47:46.830]  And if you practice good OPSEC, they can see that you're doing something suspicious.
[47:47.490 --> 47:49.730]  I mean, are we just hosed, or...?
[47:51.110 --> 47:51.930]  Well, maybe?
[47:52.210 --> 47:53.270]  Like, yes.
[47:53.810 --> 47:54.510]  Yes, we are.
[47:54.650 --> 48:04.910]  I mean, of course, the XKEYSCORE revelation kind of underscores that, that, you know, of course, like, we are looked at as very suspicious.
[48:05.150 --> 48:08.210]  I remember, like, seeing a Tor t-shirt that, like, is really kind of awesome.
[48:08.270 --> 48:10.110]  It says Tor-erist on it, you know?
[48:10.330 --> 48:17.770]  Like, just by having the curiosity to pursue a lot of these things does, yeah, mark you as suspicious.
[48:18.470 --> 48:28.650]  You know, I mean, of course, like, we've seen miraculous things happen with, like, for instance, the HTTPS campaign, where all of a sudden it became cool for these things to happen.
[48:28.890 --> 48:37.050]  Like, the work done by groups like Fight for the Future, you know, they popularized these things to the point where, like, we can normalize it.
[48:37.130 --> 48:38.190]  We can all normalize it.
[48:38.610 --> 48:46.990]  And also, like, there is a little bit of, like, security through obscurity in, like, which I don't believe in, of course, like, cross the board, but only theoretically.
[48:47.570 --> 48:59.550]  So, like, you know, like, everything on this stuff is, like, you know, code is maintained by small developers or, like, groups of people or, like, you know, students who had, like, this great idea, and they want to throw up a proof of concept.
[48:59.750 --> 49:03.290]  And, like, the very, very... I mean, it's just the idea that we have so much choice.
[49:04.150 --> 49:05.530]  Some tools work better than others.
[49:05.710 --> 49:06.750]  Everything needs to be audited.
[49:07.130 --> 49:08.310]  Always trust your dealer.
[49:08.310 --> 49:14.210]  But there is a, like, there's a fuzziness factor that comes with variety.
[49:14.670 --> 49:17.050]  Yeah, we often talked about that being our closing remark.
[49:17.050 --> 49:18.210]  Always trust your dealer.
[49:18.910 --> 49:19.770]  At the end of this talk.
[49:20.010 --> 49:31.310]  And actually, one thing that would be beneficial to underscore is, like, you don't necessarily have to contribute to an open-source project, like, for mobile or something like that.
[49:31.310 --> 49:43.030]  But it actually is a really, really... It would be a good idea for everyone to just kind of learn how to get themselves on GitHub, pull down something from source, and, like, maybe, you know, like, boot it onto your phone.
[49:43.350 --> 49:53.250]  You know, like, that's... And that will kind of cover your tracks if, like, you know, people are analyzing, like, you know, the apps that you're downloading from the Play Store, for instance.
[49:53.390 --> 49:56.110]  Or, of course, like, the Play Store is, you know, down for whatever reason.
[49:57.010 --> 49:59.150]  Many, many thanks for all of your great work.
[49:59.290 --> 50:04.290]  On passwords, a lot of people use the first letter of a phrase.
[50:04.570 --> 50:07.490]  And perhaps your mnemonic would be flag, dollar sign.
[50:07.650 --> 50:13.210]  And it would be the first letter of, oh, say, can you see, dollar sign, by the dawn's early light, dollar sign.
[50:13.350 --> 50:16.530]  And then you might have sign on to jabber one.
[50:17.010 --> 50:23.430]  And so all your mnemonic is just flag, dollar sign, or shining with a funny...
[50:23.430 --> 50:25.710]  if they allow you to use the special characters.
[50:26.990 --> 50:31.030]  But the problem is a lot of websites don't let you use special characters and passwords.
[50:31.510 --> 50:36.110]  And it was something unique to you that no one else would guess if they saw a flag, dollar sign.
[50:36.270 --> 50:36.830]  So that may help.
[50:37.150 --> 50:47.090]  My question is, is there anything specific that you recommend about key loggers or devices, any sort of software that might have been put on a PC to track all of your activity?
[50:48.910 --> 50:49.690]  Nail polish?
[50:49.930 --> 50:49.990]  No.
[50:49.990 --> 50:50.470]  Yeah.
[50:51.370 --> 50:52.510]  Glitter nail polish.
[50:52.670 --> 50:54.150]  Glitter nail polish.
[50:54.190 --> 51:04.170]  I didn't used to do this, but I keep my devices under my physical control, or the physical control of someone I trust a lot.
[51:04.530 --> 51:13.790]  And I have tamper-evident stuff that would make it harder to open it up and me not notice.
[51:13.790 --> 51:23.950]  But, you know, the truth is, how often am I going to pull out, you know, turn the device over and look at the nail polish pattern and compare it to the photo I took of it?
[51:24.070 --> 51:24.110]  The creepy map?
[51:24.110 --> 51:25.370]  The creepy map that does that?
[51:26.410 --> 51:27.470]  It's a very easy map.
[51:27.470 --> 51:36.250]  You get yourself into a rabbit hole where you spend, you know, six of your hours every day trying to curate your security stuff.
[51:36.350 --> 51:38.050]  And sleeping with a laptop is rough.
[51:39.150 --> 51:39.770]  Thank you.
[51:43.030 --> 51:44.470]  Oh, bummer.
[51:44.730 --> 51:45.350]  Thank you very much.
[51:45.390 --> 51:47.270]  Send us questions on the GitHub repo.
[51:47.490 --> 51:47.530]  Yeah.
[51:48.390 --> 51:48.670]  Please.
[51:49.070 --> 51:50.810]  Thank you for listening.
[51:57.060 --> 51:59.220]  I imagine people will have a lot of questions.
[51:59.420 --> 52:01.300]  We do need to keep things moving through the tracks.
[52:01.300 --> 52:07.900]  So, if you have questions for the speakers, please walk with them out to their next place.
[52:08.600 --> 52:09.240]  I'm speaking...
[52:09.240 --> 52:10.820]  I actually should just be speaking with you folks here.
[52:11.860 --> 52:13.880]  If you have questions, please walk with them.
[52:14.060 --> 52:14.460]  I'll go.
[52:41.880 --> 52:42.640]  Give me one second.
[53:12.060 --> 53:13.460]  I'll give you a couple of time.
[53:15.720 --> 53:17.120]  Bye.
[53:49.500 --> 53:50.900]  Bye.
[55:45.960 --> 55:46.220]  Bye.
[55:46.220 --> 55:46.580]  Bye.
[55:59.260 --> 55:59.560]  Bye.
[57:34.650 --> 57:37.450]  And this is the second one.