Open Source and Arms Control

Aaron Stein 0:00

Hello and welcome to the Arms Control Wonk podcast. My name is Aaron Stein. I'm a contributor to the Arms Control Wonk website as well as an associate fellow at the Royal United Services Institute and a PhD candidate at King's College London.

Jeffrey Lewis 0:13

And I'm Jeffrey Lewis. I'm the founding publisher of Arms Control Wonk and a scholar at the Monterey Institute.

Aaron 0:20

Today we're going to be talking about your recent trip to Wilton Park. This stately manor right outside London where they hold conferences, mostly on non-proliferation and arms control.

Jeffrey 0:34

It's Downton Abbey with a bar.

Aaron 0:37

It is Downton Abbey with a bar. I mean, I've been there once myself for a more Middle East nonproliferation conference. And I'd say it's one of the most stunning places to hold a conference. Was this your first trip or have you been there before?

Jeffrey 0:50

It was. It was. I had never been before.

Aaron 0:52

So apparently they have a lot more.

Jeffrey 0:54

Right. We're in the nonproliferation field. We know about the nonproliferation piece. I guess they have a bunch more conferences in a bunch of different areas. And so everybody was talking about the fact that Angelina Jolie had been there not so long ago.

Aaron 1:06

Oh, really? Did she stay in the rooms? They're a little small.

Jeffrey 1:11

I just imagined her sort of helicoptering in and out. Actually, you know, I'm joking, but the CEO, who now they call him the CEO, the original title was Warden, which is much better, said that she was super interested and took it really seriously.

Aaron 1:31

I've actually heard about it, that she's not just all fluff, that there's actually some substance going on behind all of her humanitarian work. But today we're not going to be talking about Angelina Jolie. The conference you were at mainly focused on open source and arms control verification. Is that correct? I was following some tweets online. Yeah, I think it's a really...

Fantastic new field. I mean, I use it a lot myself in my own research. I use it in my dissertation research. And I know that you over at Monterey... I mean, I would say that Monterey is leading the charge on this for arms control. Would you agree with that? I mean, self-promotion for where you work, where I went to school?

Jeffrey 2:12

I'd be too shy to say that, but I'm delighted when other people say it. That's certainly what we're hoping to do.

Aaron 2:19

Yeah, I mean, one thing... I mean, let's start with some of the work you've done, because the first piece that I remember reading about you that I thought was just amazing was the way that you used open source to find North Korean transporter erector launcher vehicles, or at least where they're made.

Jeffrey 2:39

Yeah, that was one of the case studies I gave in my talk at Wilton Park.

Aaron 2:43

I mean, perhaps you could just describe it a little bit, because I actually used that for an intro into how you can use open source to find corruption in Turkey.

Jeffrey 2:51

Oh, I didn't know you used that.

Aaron 2:53

I did. It just is the first sentence.

Jeffrey 2:56

All right. Well, how to do that quickly. So there was this film. Oh, I'll start this way. When the North Koreans paraded the new ICBM through Pyongyang, it was on these trucks. And the bloggers noticed that the trucks were perfect matches for Chinese trucks. And the Chinese kind of were caught off guard, right, because exporting the trucks that carry the missiles is a violation of sanctions.

And the Chinese said, oh, well, we just sent them the trucks. We didn't send them the erector, the part that lifts up the missile in order to launch it. And so we wanted to answer the question of whether that was true or not. And by total coincidence, there was a one of these films celebrating Kim Jong Il that had like 20 seconds of video of a factory where it looked like that was the place where they added the erectors.

To make a very long story short, we were able to model the inside of the building, which was really unusual. And once we had the inside, we could model the outside. And then we were able to find it in a satellite photograph. And it turns out they had modified the building in order to accommodate the larger ICBM launcher at the same time that they bought the trucks.

So it was pretty decent circumstantial evidence that that's indeed where they make the The trucks that carry their missiles, and by the way, something we don't often point out, there's a vehicle chassis factory nearby.

Aaron 4:26

Well, when I read that piece, it actually gave me an idea. I mean, what was going on in Turkey at the time was there were all these audio tapes that were being released online, open source, purportedly of the prime minister or his son, involved in shady business deals that implied corruption. And one of them was the building of very nice, they call them villas in Turkish, we call them sort of like houses, houses.

In a very nice secluded bay right on the Aegean Sea. And as a way of explaining away these buildings, one of the prime minister's close political confidants said, no, no, I've owned this land for 30 years and these villas have been there all along. That gave me the idea to find the site on Google Earth. And then you realize, no, no, they haven't been there all along.

They started to be built in 2012 and were completed relatively recently. So it's a way that you can catch governments in lies or at least when they're stretching the truth.

Jeffrey 5:20

That's amazing, by the way, that he thought he could get away with that lie.

Aaron 5:24

That's the point. This is the point that I think that we should talk about is that the way – it's getting more difficult to lie or to stretch the truth a little bit if we want to be nice to these people. To stretch the truth – About things that you could pick out using Google Earth. I mean, things that aren't very difficult to find.

I mean, and this has applications for arms control, for following conflicts, I mean, the Syrian conflict, or seeing if a government's corrupt or not.

Jeffrey 5:51

Yeah, so one of the other case studies we did that I talked about at Wilton Park involved this Burmese entity called the Directorate for Defense Industries, which is run by a guy named Thane Taye. The closest Burmese pronunciation I could do. Apologies to people who speak the language. And he runs all the cooperation with North Korea. And there are pictures of him.

I guess he's a fluent Korean speaker. And so there are pictures of him at all these meetings in North Korea. And the Burmese were supposed to cut off cooperation with North Korea as part of the engagement process that the Obama administration started in 2009. And this guy's been sanctioned. And so one of the things we've been doing Is we've been tracking the expansion of the sites that he controls in Burma since 2009, right?

And including sites where we think there are North Koreans at, you can see that these places are getting bigger, not smaller. And the best part is they always build them the same. They have all the same signatures, so it's super easy to pick them out.

Aaron 6:57

Right. It speaks to the stuff that I was talking about earlier. It's a way in which to keep governments honest. And I don't think they're getting fast enough at learning these lessons about the way in which information is being transmitted via social media or easily accessible tools like satellite photography. And I'm thinking now about the conflict in Syria.

And it was something that we were talking about off mic before we started here, but it was just the way in which perhaps government intel agencies, when they're making cases for preferred policies, are falling behind the times of people who are able to quickly pull out information that's probably just as good as the information they have about massive events.

And I'm thinking about the chemical weapons use and all the analysts that came out from just open source guys about what happened on that day.

Jeffrey 7:51

Yeah. You know, I wrote something for foreign policy at the time making precisely this case that when I looked at the, so when Syria used, when the Syrian government used chemical weapons in August, it was August, right? Yeah. I was struck by how the U S U K and French governments made their case that something had happened.

And that they needed to do something about it because all of the dossiers they released were mostly words.

Aaron 8:25

Yeah.

Jeffrey 8:25

And they were written in this sort of authoritative voice of God style, right? We cannot tell, you know, based on sources X happened. And they, I think they went over like a lead balloon precisely because people had gotten so used to getting really detailed open source information for free. And the case I'm thinking of is Elliot Higgins who went by the, The name Brown Moses, who was taking all of these YouTube videos and had figured out what the munitions were and just really created a case that 20 years ago you would have thought only an intelligence community could make.

He was doing it online, open source, for free with no special expertise other than being really smart. And I think the juxtaposition of these Lame, text-heavy government dossiers that said, trust us, with the kind of information that was available publicly, even though they said the same thing, really undermined the credibility of the governments, because it's like, well, you know.

Can't you do better than this unemployed statistician hanging out in the UK?

Aaron 9:42

Right. No, I mean I will say that the French report was quite good. There was a lot of great information in it because they released – It was better.

Jeffrey 9:48

It was the best of the three.

Aaron 9:49

Right, because it had some stuff about the stockpile that had been released publicly that was only hinted at in other things. But yeah, no, I think Elliot Higgins deserves the most and most credit in this regard.

Jeffrey 10:02

Aaron, I have to interrupt you. Do you know what the low point for me was? What? Was when the government declassified a dossier the u.s government i can't call it shouldn't call the dossier i've been in the uk too long the u.s government declassified an intelligence report on the syrian stockpile it was from like 1984 i i to be honest i i missed this and i should it's like really guys i think you could probably do better than that that's funny no i mean one thing that i'm seeing is as somebody who follows turkey primarily for my my day job

Aaron 10:32

You can do a lot of stuff with these YouTube videos. It just takes time and a lot of effort. And I think Elliot Higgins really led the way. And he certainly inspired me to do a lot of work on it.

Jeffrey 10:41

Yeah, I was going to say, I mean, the work he did on finding the checkpoint that showed that the rockets were probably in range of the targets was good. But it's not as good as your truck work.

Aaron 10:56

Well, thank you for that.

Jeffrey 10:57

Yeah, I mean, I guess I'll tell people about your truck work. It's pretty amazing.

Aaron 11:01

Yeah, you know, for those who don't follow what's going on in Turkey and Syria, there's this little enclave of Turkish territory like 20 kilometers from the inside of Syria on the eastern bank of the Euphrates River. And there's a tomb where one of the old Ottoman shahs is buried. And this is a very big deal for Turkey.

It's sort of the only enclave of Turkish territory outside of its national borders. And a couple months ago, it was being threatened by being overrun by these guys, ISIS, the Islamic State of Iraq and Al-Sham.

Jeffrey 11:36

And... We've been overrunning a lot of things.

Aaron 11:38

We've been overrunning it. They haven't overrun it yet. And that's another thing you can follow on open source. Is... They had to resupply it. And they had been resupplying it with helicopters, I've been told, by people who know these types of things. But they decided to go with convoys this time around. I don't know why, but my guess is there's a fear about MANPADS.

And they had to enter through one border gate that's controlled by a group called the PYD, which is a group that's linked to the PKK, which Turkey considers a terror group, as does the U.S., And as they started going down the convoy, you started seeing tweets from ISIS guys saying that they had taken Turkish soldiers hostage. So one must take a lot of what these ISIS guys are saying on Twitter with a grain of salt.

So I tried to do my own Elliot Higgins and put together stuff on the YouTube videos because YouTube videos started coming out of the convoy.

Jeffrey 12:42

So you could figure out where they were.

Aaron 12:44

So you could figure out where they were. And then what I quickly found out is that the ISIS guys were telling the truth. They came in through Kurdish-controlled territory, but they crossed the bridge over the Euphrates and then went back up north through another border gate that's controlled by ISIS.

Jeffrey 12:56

So they went in at one point, but then they didn't retrace their steps to go back. They went through ISIS-controlled territory and returned to Turkey via a totally different and presumably way more dangerous route. Exactly. So in my mind, I'm saying, God, if this were me, I would typically just turn the truck around. You see the little black flag waving, and you're like, oh, hey, okay, goodbye.

Aaron 13:18

Yeah, I mean, even though the PYD are – Not on Turkey's Christmas card list, they certainly have better cooperation than these guys with ISIS.

Jeffrey 13:28

It was a terrible mixed metaphor, given that I'm certain Prime Minister Erdogan does not send out Christmas cards.

Aaron 13:33

No, no, no, he probably doesn't. But anyway, so the point is, so what I was able to do, I said, well, could ISIS have taken these guys prisoner, or are they sort of spouting off a little propaganda? And so what I tried to do is I used the open source videos to track them along the way. And so I used Markers that you can see in the background of the videos, like mountains.

And I would look at a topographical map and say, well, this is the road they came in on. That's probably the mountain. So they're here. And then there's a great website called suncalc.net. And so you could track the trace of the sun. And I would look at the shadows to get an estimate of what time it was.

I mean, I can't say it was perfect, but I think I can get within an hour. What you can clearly see is that ISIS guys probably didn't take these guys hostage because the time doesn't work out. There is a gap of time there, but what ultimately swayed me in the fact that the story wasn't true was that ISIS media was saying that they had taken some of the Turkish trucks hostage.

And that was pretty easy to count. You could just count the trucks going in and out, and they matched. So that's the way I was able to basically disprove ISIS. But with that said, it still leaves a lot of questions about why they didn't just turn around the trucks and go home.

Jeffrey 14:49

So it's like a mystery why they chose to drive through this incredibly hostile, dangerous ISIS area.

Aaron 14:55

Right. So you can get a sense that something happened. The ISIS guys clearly were reporting something. The Turks were very hush-hush about it. Probably not prisoners, but maybe they came in contact. They certainly did come in contact because ISIS put out videos, and I was able to find where ISIS shot the videos from markers.

Jeffrey 15:13

That's really cool.

Aaron 15:14

And so they probably had to have a chat somewhere along the way. And I don't know about you. I don't like to have a chat with these people.

Jeffrey 15:23

No. I guess that would be T.

Aaron 15:25

Yeah, maybe they had a little tea. So, I mean, I have my own thought.

Jeffrey 15:28

Tea with ISIS?

Aaron 15:29

Tea with ISIS, probably in a city called Monbij to allow for safe transfer of these soldiers. So something happened, and you were able to figure it out.

Jeffrey 15:42

I mean, what's amazing to me about this work and what really came up at our meeting at Wilton Park, and the rule at Wilton Park is you can use the information, but you can't attribute it to particular people. But I think I can safely say People who work on security issues full-time inside governments aren't necessarily aware of how much information is now available publicly.

And this idea that you can use YouTube videos... To track a convoy and then more or less create a very strong timeline that eliminates some possibilities and raises others. My experience is when you explain this kind of stuff to people, the third case I used was we found a, should I say, a secret Chinese underground nuclear reactor that never operated.

Beyond the one the Chinese have admitted to. I'm going to outline it actually in my Adelphi book.

Aaron 16:52

Okay.

Jeffrey 16:53

So it's going to be public sooner or later. It's a plug. Yeah, it's a plug for the Adelphi book. Very British kind of... I'm doing a lot of British stuff.

Aaron 17:02

Yeah, I work for a British institution.

Jeffrey 17:04

You what?

Aaron 17:05

I work for a British institution.

Jeffrey 17:06

Yeah. So when you confront government officials with this kind of information, they are really not emotionally prepared for it. I mean, they really think... That they have all these secrets. And when you say like, Oh no, no, no, I know X because, and then you kind of put that information out. They kind of go through like, it's like stages of grieving.

Aaron 17:39

Yeah.

Jeffrey 17:39

You know, and, and, and you know, there's a wild, my favorite stage, which I think is the one that's worth talking about is the which I think you see a little bit in Turkey, right? But it's the Well, we'll just shut Twitter down, or we'll just, you know, we'll just stop doing whatever, as though, like, the internet is going to go away.

Aaron 18:00

Yeah, no, yeah, the Turkish Minister of Intelligence, his name is Hakan Fadan, wrote his PhD dissertation, in English, by the way, so everybody can read it, about social media and new media for arms control verification. And I sort of view this also along this paradigm of open source. When Turkey shut down social media, I just pulled up his dissertation and started reading it.

And he has multiple passages in there about... I guess I should say before he did his PhD dissertation, he was in the army. So he's been in government for a very long time. He writes about the threat to government losing its... Its monopoly on the dissemination of information and how terrorist groups can use this stuff to undercut the government's message.

And that's what was going on. If you were following anything in Turkey about all these phone calls that were being recorded online, it wasn't a terrorist group per se, but it was a group of loosely affiliated businessmen all operating under the, I guess, tutelage of an imam that lives in Pennsylvania. The story... Turkey could be an episode of 24 every day of the week.

But the point is that their reaction to this was they treated this outside group, the Gülen movement, as a terror group, and then they just said, okay, we're just going to shut Twitter down. So people cannot listen to these telephone calls. And I think his dissertation provided the blueprint for what the government's policy reactions were. And it was all written down in his own words on pieces of paper that you can just Google if you know where he went to university and get his dissertation.

Jeffrey 19:49

Gee, Aaron, that sounds like that'd be a nice blog post.

Aaron 19:52

I'm happy to – I actually did write a blog post but on my own blog, Turkey Walk, in case anybody wants to read it. But I can link it. We'll link it in a box. That doesn't help me, Aaron. You can read it.

Jeffrey 20:02

That doesn't help me in my hunt for content.

Aaron 20:05

I know something else. I am writing this for the Arms Control Wonk blog but it's been slow going because I want to make sure I'm absolutely correct. Again, I view this on the same paradigm as this man, this Danilenko man, who is alleged to have helped Iran with the production of nanodiamonds, which we joked about in another podcast, wrote numerous open source articles and has numerous patents in the Russian database, which are also available online.

It's sort of a wild beast to go in there into the old Soviet archives, but you can find his patents online. And in it are a lot of stuff about detonation chambers. And, you know, I hesitate to say it because I haven't quite... But, you know, you can see some discrepancies between the way he ideally would like to produce nanodiamonds, the size of the detonation chamber, versus what we know about the size of the alleged detonation chamber at Barchin.

You know, it's three times as large. And Donalenko writes about why such a large detonation chamber isn't actually very good for nanodiamond production in his own work. So this is a to-be-coming on Arms Control Walk. It's just moving slowly because I'm waiting for one last piece of information. But I view this all in the same timeline. Everything's so easy to find now online that it's hard to keep secrets.

Jeffrey 21:32

Yeah, absolutely. So we did a Podcast, actually, for Wilton Park. I'm the guy who was on my panel, Carl Miller, from something called Demos. And we spent a lot of time just talking about it. He's more of the kind of big data approach. Can you mine massive data sets in order to find information? But it was, yeah, it's super clear that...

Well, I would say it this way. When I was a research assistant at CSIS, I worked on this project on the information revolution in international security. And this guy wrote this totally crazy paper that I didn't get at the time. But he said this really insightful thing, which is you're not going to be able to get rid of the Internet by trying to block little bits of it here and there any more than you could stop the coming of the railroad by tearing up some track.

And I think he had this idea, which I think is right, that so much of modern society depends on this infrastructure that we've created and this way of living and this constant churning of information that it's really hard to shut it off. If you want to shut it off, you've got to shut it off like North Korea shuts it off.

Right. And even then, it's not perfect. No, I mean – People really live online now.

Aaron 23:00

Yeah, no, and in this dissertation I was just talking about by this – by Turkey's head of the intelligence. He says this exact thing, not in those exact words. I'm paraphrasing. We can't stop this. So government has to compete is essentially what he was saying. And even then, when he writes this in his dissertation, what I think was his policy guidance to the prime minister, because they met the night that these tapes started coming out, like two hours after they met, he was the first man that the prime minister met with.

Even then governments act rationally, you know, act sort of not in their own self-interest. You know, the smart people realize that they can't stop the flow of information, but they still make the decision to try and stop it. And I think that's something that we're going to see a lot of moving forward in the future.

Jeffrey 23:47

Yeah, I teased Selstrom about this because, you know, my presentation had been about North Korea and China and Burma, and there was a little bit of like, you know, Whatever, because I run our East Asia program, so I did a lot of East Asia stuff. But it was, you know, it's not just governments in East Asia. When Selstrom handed over his report to Ban Ki-moon on the chemical attack in Syria, there were a series of high resolution photographs taken and he had tipped up the report as he handed it to him.

And so reporters got a clean shot of the front page. So it was embargoed. But news agencies were immediately able to report on the findings because he had tipped it up. And I mentioned that story, you know, just sort of teasing him about it lightly. And he had a very good sense of humor about it. I mean, it wasn't like a thing.

But apparently this has now happened in the UK a number of times, which is, you know, people carrying documents get photographed and pretty quickly that information gets out. So it's just... You can make just the slightest mistake. And the thing is, because information is so cheap to move around now, You know, the slightest mistake is immediately populated everywhere.

Aaron 25:03

No, and I actually just saw a presentation about this at RUCI in the UK. This is a UK-themed podcast about violations of export controls.

Jeffrey 25:11

It's a brilliant podcast, Aaron. It's brilliant.

Aaron 25:13

This is a brilliant podcast. It's violations of export controls by engineers on Facebook who, like, post pictures, like selfies of them, in front of sensitive material, like, say, the aeromissile. I'm not kidding. He didn't name the country, but he slipped and said the Aeromissile, so you can guess which country he's talking about.

Jeffrey 25:31

That's great, because my colleague Amy Smithson refers to all the open source stuff that Brown Moses has done, a lot of which is put up by the people who are committing the atrocities themselves as war crimes in the age of selfies.

Aaron 25:45

That's right. So they posted selfies of themselves in front of the Aeromissile, apparently with something open to where embargoed components, U.S.-made... Were then spread across Facebook.

Jeffrey 25:57

It's just spectacular.

Aaron 25:59

It actually has legal ramifications because he's an export control advisor for a large company. I won't say it. It's actually causing legal problems in terms of export control violations on U.S. companies of taking selfies in front of things that you're not supposed to take selfies in front of and posting them on Facebook. It's good for us.

Jeffrey 26:21

Absolutely.

Aaron 26:23

Yeah, well, Bram, perhaps we should just wrap up there with maybe what projects we're working on now and how we're using open source. Yeah, sure. And then, you know, we can wrap it up. Okay, well, what are you up to? Well, you know, I told you, I mean, I'm working on that larger post for Arms Control Walk about the detonation chambers and trying to weave Danilenko together.

And I'm also... Well, I guess what I've been doing is that I've been keeping tabs on the prime minister's residences down in a very nice bay. But no, I'm just kidding. More seriously, you can see in Iran, if you look at older photographs, like the size of some of their nuclear facilities, and you can count buildings and then You can read then, say, open source documents about them, like archival documents from the U.S. archives.

Like the Isfahan Nuclear Center had 15 buildings, right? And then its main building was something like 30,000 square feet. It's a very large building. So I'm trying to take that. Oh, and it was designed by the French, okay? So that's the background. So what I'm trying to do is compare that building with anything we know about... The Israeli nuclear program and to see if the building matches, see if the two buildings match in terms of size.

And then I'm just counting how many buildings are on site to see how fast the program evolved and see if we can't find a date where like there's just a big rapid expansion and see if that date doesn't correspond with some other stuff we know about the program.

Jeffrey 28:00

Oh, that's great. I've got, I've got a, Bunch of work on sites in China that came out of writing the Adelphi paper watching them expand their civilian centrifuge capability to make enriched uranium for nuclear fuel but also there are now thermal images available so we've been playing around looking at stuff that's hot to see if facilities are still active So that's one whole set of stuff.

A bunch of stuff on North Korea. I've got a piece coming out in 38 North. And actually other people have done precisely the same analysis because it wasn't really all that complicated. But the North Koreans released an hour-long movie of Kim Jong-un watching artillery barrages, often female unit artillery barrages. Oh, Jesus. And at the 49-minute mark, there is a second-and-a-half clip of a missile launch.

And it turns out it's a sea-launched... A cruise missile. Yeah, a cruise missile.

Aaron 29:05

I've seen this stuff.

Jeffrey 29:07

And everybody expected that the cruise missile would be Chinese or perhaps Iranian origin. And it's a dead ringer for a Russian cruise missile called the KH-35, which just opens up all these interesting possibilities. But you really have to pause it, right? Because it is literally a second and a half. And it's so easy to miss because I got to tell you, after 49 hours of North Korean artillery barrages, it's pretty boring.

And then Burma. You know, we want to do some – well, I should say there are two more things. There's Burma because we keep finding new sites. And I guess I talked about it on the podcast, so it's not such a big deal. Also mining Russian Facebook, the Russian equivalent of Facebook.

Aaron 29:51

I guess Israeli Facebook, from what I understand now, might have pictures of the arrow.

Jeffrey 29:58

I'm totally going to spend a little bit of time on that.

Aaron 30:01

Yeah, and I guess I'll add too, I'm taking YouTube clips of the conflict in Syria where they're using these tow missiles, these missiles that were probably signed off on to be sent to the conflict by the U.S., probably provided by Saudi, Through this place called Hatay in Turkey, and I'm trying to actually use geographic markers to find where they're actually being used in the conflict.

It's proving very difficult, and it may not work. But then try and drop those thumbtacks on Google Maps onto where they're being used to see if we can't, like, the breadcrumbs don't lead back to a specific border gate on the Turkish border. So I'm trying to do that, but it's proving to be extremely difficult. But yeah, those are some of the stuff I'm working on, and hopefully it'll come out in the next couple of months.

Jeffrey 30:48

Awesome.

Aaron 30:48

Yeah. Well, thanks for talking again this week, and it's been good to talk about things that end on a positive note.

Jeffrey 30:56

Yeah, I know. Two in a row where we've been all cheery.

Aaron 30:59

Let's keep this going. Absolutely. So thanks for talking, and thanks for listening. All right. Bye. Bye-bye.