Angry Planet - Bellingcat Talks Tankers, Deepfakes, and State Propaganda

Episode Date: June 28, 2019

Misinformation. Grainy and low resolution images from government sources. Signals intelligence. Satellite imagery. Photographs and social media posts from those nearby. Online databases of vast and public knowledge. These, and more, are the tools by which Eliot Higgins and Bellingcat suss out truth in a complicated world.On June 13, something attacked two oil tankers in the Gulf of Oman. The United States blamed Iran, producing footage it claimed linked Iran’s Revolutionary Guard to the incident. Iran denied responsibility and people aboard the tanker say the story the US is telling doesn’t add up.So is this a clever sleight of hand from Iran, a Gulf of Tonkin style operation meant to draw the US into war, or something else entirely. Here to help us figure that out, and more importantly how he came to the conclusion, is Eliot Higgins of Bellingcat.You can listen to War College on iTunes, Stitcher, Google Play or follow our RSS directly. Our website is warcollegepodcast.com. You can reach us on our Facebook page: https://www.facebook.com/warcollegepodcast/; and on Twitter: @War_College.Support this show http://supporter.acast.com/warcollege. Hosted on Acast. See acast.com/privacy for more information.

Love this podcast? Support this show through the ACAST supporter feature. It's up to you how much you give, and there's no regular commitment. Just click the link in the show description to support now. And in fact, the video was from a computer game, and the Russian MOD had actually used a screenshot from a computer game as its irrefutable evidence. You're listening to War College, a weekly podcast that brings you the stories from behind the front lines.

here are your hosts Hello, welcome to War College. I'm your host, Matthew Galt. Producer Kevin Nodell is still in the Middle East, but we will hear from him soon. Misinformation, grainy and low-resolution images from government sources, signals intelligence, Facebook posts, satellite imagery, YouTube videos, online databases of vast and public knowledge. These and more are the tools by which Elliot Higgins and Belling Kat suss out the truth in a complicated world. On June 13th, something attacked two oil tankers in the Gulf of Amman.

The United States blamed Iran, producing footage. It claimed linked Iran's Revolutionary Guard to the incident. Iran denied responsibility, and people aboard the tankers say the story the U.S. is telling doesn't add up. So is this the clever sleight of hand from Iran, a Gulf of Tonkin-style operation meant to draw the U.S. into war, or something else entirely? Here to help us figure that out, and more importantly, how we can't. came to this conclusion is Elliot Higgins of Bellingcat. Elliot, thank you so much for joining us. Hello.

So very briefly, for those who might not know, can you briefly describe what Bellingcat is and what its mission is? So Bellingat is probably best known for the online open source investigations we do. We investigate various kinds of incidents, mainly focused on conflicts, but also broader than that. But we do that by using publicly available information, looking at, for example, social media posts on a range of platforms and also using other resources like Google Earth, Google Street View and different kinds of searches. But it's really about trying to connect to as much digital information

as possible, you know, using the internet, you know, to its full extent, to explore what happened in different areas. All right. So June 13th, we hear about these tankers. What's the first thing that Bellingcat does? So the first question is, you know, are these tankers actually where their claim to be. And we can actually use websites now, like in marine traffic, that actually tracks the transponders on these vessels. So you simply search for the names of the vessels that claim to have been attacked, and it shows you where they were. They did show that they were in the Gulf of Armat quite close to each other. And then pretty quickly after information starts to come out, the US government then releases images of what it claims

are a hole blown in the site of the hole by Olympat Mine. And what did you make of these? and how did you process them? Well, I mean, what we were looking at, we had this blurry footage of this vessel that was claimed to be Iranian that was supposedly taking something off the side of the vessel. Now, this was very blurry footage. It was hard to see what was being taken off,

even if there was something being taken off. The vessel, we were told it was Iranian, but how do we know it's Iranian? So, kind of various people online, including myself, started to do. digging through video footage and photographs posted by the Iranians themselves of the vessels they use. And it was possible to find a match to the same kind of vessel that was featured in this video. So by doing that, we can start kind of establishing that whether or not the story that's being

told by the Americans is consistent with the evidence they're presenting. Now, I say consistent because what they were actually presenting was not necessarily proof of what they were claiming. We could see in the video that it kind of looked like someone was removing something. from the side of the vessel, but how do we know it's Olympic mine as claimed? We don't have a clear view of it. So we are trying to establish as much as we can about it, but what I think is very interesting in the way that states, and, you know, including the US, are presenting information about their claims is they could be a lot more transparent. Why is it up to the open source

community to figure out whether or not the boat featured really is the boat they're claiming? Why didn't they present their own evidence making this case? And this is something I come back to time and time again when nation states are trying to make these claims that we have this vast amount of open source information that can be used to verify the claims they're making and they're just not bothering doing it and I think there is this kind of perception among people who are kind of very aware of what you can do with open source information that they are doing this and it creates this kind of dissonance where kind of doubt about the claims kind of creeps in it's kind of the extraordinary claims require extraordinary evidence in a world where

everyone's filming everything all the time and everything is tracked and monitored, you better be able to back up what you're claiming. That's right, and especially when you're dealing with Donald Trump and John Bolton, who have a loose ethics when it comes to the truth. I think most people want that extra step. They want clear evidence. I mean, something as simple as the video starts at what appears to be the moment the mind is being removed. Why don't we have 10 seconds beforehand when that obviously?

would be a lot clear on the whole. It just adds questions in the minds of the people who are already quite cynical and dismissive to these claims. Now, there's always going to be an audience that is going to be dismissive no matter what the evidence is, who are convinced that everything's a false flag. But they can do a lot more to make their case and they're failing to do so at the moment. We've had a bit more information. There was a clearer photograph of the vessel that allowed us to confirm the identity of it. There was an image showing a close-up of the side of the vessel from where the object is removed some sort of disc-shaped of broken objects attached to it,

which appeared to be in a ring of similarly shut-sized marks on the hull, which suggested that could be a kind of magnetic kind of clamp, but it still didn't tell us a type of munition. Now, what would have been useful there, again, is if the US had an image of an Iranian limpet mine that had a clear view that matched the object, that was kind of left on the hull. So there's options for the US to make a much clearer case,

and it's something they've been failing to do. That's not to say that they're lying, but they need to make a case to the public about what's happened. What is, can you describe to the audience, what a limpid mine is and why that might have been the weapon used? So it's basically a shape charge that they can attach magnetically to the side of a vessel and then detonate and put a hole in the, the side of it. And we could see from the photographs cleared there were holes in the side of

these vessels and one of them was some fire. But the question is what made those holes? And then you have the crew who were claiming there were flying objects. They spotted just before the vessel was hit. Some people saying those were missiles being fired. We don't have any direct evidence of that, but having these kind of contradictory statements or statements that appeared like they might be contradictory makes masses more complex. And what do you think is missing from the footage? What else are we not seeing that we should be seeing that would maybe clear anything up? Well, for one thing, it would be useful to have a clear view of what if that object was on the side of the vessel.

There's a couple of photographs, but at best it looks like a black triangle, which really doesn't tell us much about the objects in question. You know, this footage that they broadcast, if we had, you know, just the moment before they approach the vessel, we'd have another angle on this object. It would help make it clearer. and just presenting more evidence that the remains of the object that was removed from the vessel and photograph was what was claimed to be, you know, the clamp on the limpid mine. Another thing I think is very is fascinating and important understand about the work you do is that anyone can do this, right? The tools exist for all of us to be our own investigators, correct? Yeah, absolutely. I mean, it's like I mentioned before, we have the marine traffic website that allowed you to track the vessels. we have satellite imagery. After the vessels were hit, there was actually satellite imagery

showing the oil that had leaked into the sea around it. So it allows us to kind of cross-reference lots of information that helps us confirm some of the claims being made. You can look at the vessels that were featured, the Iranian vessel, and find the same vessel in every imagery. And this is all publicly available information. And that can take you so far, but there is going to be information that we can't get, especially in a military context. But looking broadly, you know, it'd be to just this specific instance. I mean, Ballencastle an vast amount of investigations sticking through all kinds of material, just doing all kinds of things like identifying, for example, the missile launch of the shot down in MH17, where it came from, the military

brigade it came from, all the members of that military brigade from their social media profiles and who was in the conway that transported it, and just so many more details that can be found online. How does someone train themselves to do this? Because that's what you did, you taught yourself how to do this. Yeah, I mean, when I started off, I was just doing this as a hobby, and I was interested in videos from Syria and what they could tell us. And I didn't speak Arabic, so I was focused very much on the weapons. And there's a lot of information available online about kind of Soviet era weapons,

which were most of the weapons being used in Syria. So that was mainly just, you know, digging through stuff. If I saw a marking on a weapon, I'd Google what that marking was, and then that would lead to photographs and people discussing it. And, you know, myself, I was able to get in contact with arms experts and learning a lot from them directly. But the stuff we do is quite simple. It's like if you go to a meeting or an event and you Google the address on Google Maps, you've done an open source investigation. If you start using street, you have to have a look around the area.

That's an open source investigation. It doesn't have to be about a war crime. It can be just figuring out where a shop is. And anyone can do that. It's a simple little thing. Right. everyone has the tools already. A lot of people are already doing this stuff.

It's just about translating it to, you know, things you may not think to look at, such as, you know, tankers on fire in the Gulf of Amman. It's like having a toolbox. And, you know, it's easy to use a screw over and hammer. But you can kind of combine those in interesting ways and kind of look at the material you have and what you want to turn it into. And then you have these different solutions to get to that point. So there's this idea that.

that it's very popular right now, that we live in a post-truth world, right? That, you know, you can't trust anything you see anywhere, and it's incredibly hard to figure out, you know, what's actually going on in any one incident, especially if it's something overseas and especially if it's something to do with the military. It seems like your work and what you do stands in opposition to that. What do you think of this post-truth world?

Well, I mean, my approach has all to be looking at, you know, what can we find out from all this information that's available online? It really came to me when I was looking at what was happening in Libya. And this was just as a consumer of news, not as an investigator, seeing there was these little scraps of information that was coming from Libya that could be pieced together to get a better understanding of what was happening on the ground. And at that point, no one was really doing that. So that's what I started doing with my blog back in 2012.

And now we're looking at how we can apply this kind of investigation process to justice and accountability, working with lawyers and giving stuff to courts and seeing how that kind of information can be using courts. Right. I believe the last time we talked, you had just testified at somebody's war crime trial, correct, using YouTube footage of war crimes? It may have been the European Court of Human Rights case on MH17. And there's the Court of Justice case, Ukraine versus Russia, where we've been asked to submit evidence.

The lesson we learned there is that you need to have a good process for archiving and investigating information at the time you're doing it, because going back to stuff that you wrote three years ago and finding all the links are broken and then having to rewrite it in a certain style for a court in great detail is painful when you've already done it once. So what we're doing now with this new investigation process we've developed is we have an archiving process where as we're doing investigations online, everything we're doing is being. archived and saved. Alongside that, we can archive specific pieces of information, add lots of metadata to it and create a, basically, a giant searchable archive of verified information. The idea there is then that will be accessible to investigative bodies in the future. And it will also allow us to create a process that we can share with other organizations and have multiple archives and investigations kind of meeting the same standard.

then it's going to be kind of centrally centralized indexed, making it easily accessible to those kind of bodies that need to find this information, and don't want to spend months just investigating one incident. It's building a database that people can build on. Yeah, it's having this database, but also the triple I am on Syria that was set by the UN General Assembly is looking into the issues of how they can deal with literally millions of videos from Syria. It's a huge data management task, and that's just the kind of YouTube videos. Then you have all the videos that were filmed by people on the ground that might be on a hard drive somewhere. How do you make that information accessible when not everyone wants to share all their information with an organisation? So the idea is that we are developing this archiving software with the Syrian archive. It's part of the process that we're using, and this archive is archiving software allows it to be accessed by the index that sits on top of the archive.

Cliving Self, where I should say, allows it to be searched for information. Then we can choose to share that information with the bodies you want to find it. But it also means that we can distribute this process to other organisations, and they can also share it in the same way. And they therefore have control over what they're sharing. So it might be it's an organization that has lots of sensitive witness statements. And they don't necessarily want the whole world to know about that, but they still want it to be accessible through this centralized index.

So they can just share that they have a video filmed about an incident that to took place on a certain date at a certain location. And then if, say, the AAAM on Syria wants it, they can send a request and they get sent to them securely, rather than it being in a process where, say, an investigator, the AAA is looking into an incident, and then they have to figure out every single organization you might have a video, send them all a separate message, figure out if they can use kind of the description of the incident that all the other organizations are using, and then wait for those organizations to respond and send them an email.

something that could take weeks and weeks and weeks is instead something that can take a few mouse clicks. How do you protect a database like that? Well, this is a big issue we're facing with our Yemen project because we have concerns in how we're looking into Saudi airstrikes. So we're concerned about how Saudi Arabia will respond to our research. So, I mean, there is just kind of security of the physical server security, general cybersecurity, You know, simple stuff using the two-step verification, not being sitting and clicking on links that you should be clicking on. But there are, you know, if there's a physical attack of some sort on our server and, you know, someone tries to access our building and get into it, that, you know, we have to take physical security measures and putting everything behind locked and coded doors and those kind of ideas. Do you think it's easier or harder now for state actors to lie to the public and get away with it?

I think it's a lot harder. I mean, through the work we've been doing with MH17, we encountered a lot of Russian government propaganda about what happened. They presented a lot of evidence that we showed was fake. Or, you know, they made statements that we proved was totally untrue. I mean, the broader statement they made is that they weren't involved with MH17 being shot down. But using open source investigation, we've been able to prove that they provided a specific missile launcher that was coming from a specific. unit in a specific convoy on a specific day and across the border and was filmed in Ukraine that

went to a specific field. And this is all from publicly available information. And this is information that the official criminal investigation is basically verified at this point. And we've just seen four suspects now charged as part of the investigation. And everything they've done is consistent with the open source investigations we've been doing and making, we've made findings that they've revealed, you know, two years after we've revealed them. But what they also have is, the non-open source stuff. They're really interesting stuff like more phone intercepts, witness statements and additional details. So for Russia to keep saying that they weren't responsible for MH17 being shot down and they aren't sending military equipment across the border

is something that is clearly false. So we can demonstrate that very, very easily at this point. But how do you get people to listen? Well, I think you're always going to have an audience online who don't want to listen. They're going to see everything as a vast conspiracy. the more evidence you present just reinforces the idea that it is a vast and complex conspiracy in their view. So you can't really make those people happy or change their minds. I think what you can do is, you know, present evidence and information to a more general audience who isn't necessarily part of that kind of information ecosystem. And we'll kind of

see that information ecosystem as being something that's kind of insane and illogical once they've seeing the evidence that's out there. So it is worthwhile doing these investigations. And the great thing about open source investigation is we can present all the evidence that we're using and people can look it at themselves and make their own conclusions. Deepfakes is one of the new hot buzzwords we're talking about in the tech community. I assume you know exactly what it is, right? Yeah. So deep fakes are something that I get asked about a lot. And I kind of think of them in two ways. One is the kind of immediate social media reaction to them, where if you have a minute-long video of Barack Obama saying Donald Trump is the greatest president ever, that'll go viral and

spread across the world in a few moments. And counteracting that is difficult. On the other hand, as we're doing investigations, if we saw a video like that, we would figure out who posted it first, you know, find the original speech, the full-length version of it, watch it, you know, look at all the different aspects of it. Any actual piece of evidence is part of a network of information, and that's the network that we're exploring. So you just see it as another piece in the puzzle. Are you worried about the effect that these kinds of videos may have on the broader public discourse, people's opinions of military operations?

And we're putting our fear cart before the horse right now. But I think people that look at this stuff and understand it are really scared that these things can get out of hand. I think what you might see happening is people will lose faith in kind of videos being real, which in a way is I think how kind of social media community will respond to this. Because if we start, for example, seeing the upcoming presidential elections in the US, we start seeing kind of these videos appearing. It won't take long for them to be identified as fakes. and then any time you have a video of someone saying something extraordinary, people are going to assume it's fake, because they'll have so many examples of where someone has said something extraordinary and it's turned out to be fake. You might see that slightly differently in closed social networks, though. One thing we've seen more and more is kind of violence that's inspired by kind of conversations going on in closed social networks. We've seen this in mine and mine and mine and elsewhere in the world. And that's harder to address. because generally if you're part of a close social network, you're going to trust the other people in that network.

You aren't going to be fact-checking them so much. And they're probably going to have the same points of view as you do. So they aren't even going to be interested in checking this stuff. They'll just assume it's real. And I think that is kind of probably more of a threat with deep fakes than kind of open social media where we can kind of see these things and debate them and debunk them. Right.

This leads beautifully into my next question. How can investigators combat misinformation faster than the speed of public opinion? It might be possible. I mean, we have one interesting example where the Russian MOD posts on their social media accounts, what they claim to be irrefutable evidence of U.S. helping ISIS withdraw in Syria so that it could be redeployed elsewhere in the Middle East, as it was described. and they had four images that appeared to have been taken from a drone. Now, we looked into this because one of the images was very interesting,

because I had seen it a couple of weeks beforehand in a video posted by an Indian journalist, and that journalist had claimed it was US forces attacking a convoy, and in fact the video was from a computer game, and the Russian emoji had actually used a screenshot from a computer game as its irrefutable evidence. What happened then is because I had kind of seen this video, from this Indian journalist and then mentioned it on Twitter

saying it was from a computer game. The people who've kind of followed me on Twitter are also the kind of people who follow the Russian NOD on Twitter. So you immediately had this reaction against a Russian MOD of people saying this is from a computer game, this is nonsense, and it happened immediately.

And because of that, before even this narrative could be kind of created beyond social media that the US was doing this thing, you know, protecting ISIS forces, the narrative that was created was in fact the Russian Emorydo is using computer game photographs as irrefutable evidence

and it was widely reported it became a big news story and it was very very embarrassing. So by having this discussion about fake news looking, you know, using social networks to get in a way, create an immunity against some of this. It can be very useful and it

kind of creates a certain level of cynicism but when you have something somewhere like the Russian MOD presenting something like that as evidence it turns out to be fake, having that kind of inoculation against that kind of fake news is very useful. What is the deal with the Kremlin using video game screenshots in propaganda? It's not, that's not the first time I've heard the story or seen it. Well, it's like they claimed it was a civilian employee who did it, but it had been purposely

cropped to crop out all the data that would have showed it was from a computer game. So it was definitely something that was done on purpose. It was translated into Arabic and English and posted on Facebook. Twitter. So it wasn't just some silly, you know, employee clicking on the wrong image in a folder. It was a definite attempt to fake it. And the other three images are actually from Iraq and nothing to do with Syria and from a few years earlier. Why they do it, I think it must just be hubris. And you've got to think the people they're hiring to do these jobs are not people who are interested in, you know, the facts. They're interested in, you know, the party line,

being loyal to Russia, not about, you know, fact-checking every little thing that they're doing. So, I mean, it was an idiotic thing to do. But, I mean, in my own experience, I mean, because Bellingat has written a lot about Russia, we've had claims made about Ballenka and me personally. We had the Russian ambassador to the UK give a press conference after the scriptal investigation we did, why we identified the people who claimed to be sports nutrition salesmen, going to Salisbury to see the cathedral were actually GRU agents, and we exposed that. When he was asked about that, he started making claims that we were part of where he called

the British Deep Establishment. And he made those claims several times. And when a journalist at the end of the press conference asked him, what's your evidence? He said, well, we don't have any evidence. We have a feeling. Which really says everything about how Russia approaches these kind of things. It's all about feeling for them, not evidence. Well, and it also feels like Russian propaganda.

Andy is a hobby horse on the show. What do you make of this idea that they will intentionally release several different narratives to confuse, you know, confuse everybody and muddy the waters and kind of make everyone cynically give up? Well, you saw this quite clearly with MH17, just shortly after MH17 was shot down the Russian Ministry of Defense gave a press conference. at their press conference they presented various claims about what happened to MH17. Now what they actually did is rather

than making specific allegations, they presented evidence that if you actually looked at it, it was fairly contradictory. So one thing they presented was satellite imagery claiming Ukrainian buck missile launches has moved on July 17th. And they asked, why did

these buck missile launches move? They were very confused about it. But they also presented their evidence that a Ukrainian SU-25 was close to MH17 when it was shot down. And they were asking, why is it nearby? They presented evidence

that MH17 had dramatically changed course and asked whether or not that was Ukrainian air traffic control. So they set up all these different scenarios by asking these questions. All this evidence was fake and false and it was all lies.

It could be proved for open sources eventually. But by setting up all these different questions, then this kind of the Russian media picked up upon it. We had, you know, the kind of online community discussing it and bloggers pursuing certain things. So after that, the idea that it was a Ukrainian SU-25 that shot down MH17 kind of took hold. So then all the kind of online community discussing this were finding what they claimed was evidence that proved this. But in the end, after two years, even Russia kind of

using the same evidence actually said there was actually no aircraft near MH17. So the community around it, we've straight on to saying it as well, it's a Ukrainian. But then, even though they spent the last two years trying to find the evidence that supported it was an SD25 and constantly claiming success. And it's kind of this mindset of the kind of people who are the kind of pro-Russian MH17 troopers that if something's disproven, they just move on to the next thing that they claim proves it, even though they've been wrong 50 times in the past. It's just hard to imagine that public opinion would continue to follow them anywhere.

Yeah. But I mean, I think that's probably part of the point. Yeah, I think in a way they're kind of trying to create this. I don't know if they purposely are trying to create confusion. They just want to kind of feed, you know, plant the seeds for as many, many narratives as possible to, you know, get to the point where they want, where there's just this insane kind of debate online about this kind of stuff. But the face with MHMATI, there's a criminal investigation and a vast amount of open source evidence. So it's not going to really work in a court if they go there and start saying, oh, here's what we think might. have happened. You know, we have in a European Court of Human Rights case where, you know, they're

asking for about $3 billion for the families of the victims. And there they can't come, you know, that case has been accepted by the European Court of Human Rights. We're now waiting for the Russian response. I think the expectation is Russia is going to ask for more time, but then we're going to have the March 2020 cases in Holland on M817. And they're even more evident. is going to be released. So what's Russia going to come up with? A bunch of conspiracy theories they found on website, it's not going to work.

What do you think of legislative attempts to combat misinformation? Can government do anything about, say, like, deep fakes? I mean, it's difficult because if you start legislating against what is considered to be fake news, He basically sets up a situation where we can do it in the West and do it with the best intentions. But we've already seen countries all over the world using the term fake news to dismiss evidence. A good example of this is there's a video that was posted online in the summer of 2017 or 2018 actually. And it showed two women and two young children somewhere in Africa or assumed to be in Africa being walked off the side of the road and executed. It was a brutal video and it was quickly established that it was uniforms of the Cameroonian

army, the language match, the weapons they were using matched. And this was presented to the government of Cameroon who dismissed it as fake news, borrowing Donald Trump's term. But investigation by the BBC collaborating with Amnesty International, Banniket and others found exactly where it was filmed, the approximate date range and the identity of everyone involved. And a year later, the seven soldiers who were done. involved are now being put on trial. And this is after the government called it fake news.

But this is almost this kind of casual dismissal of evidence that's shared online because we kind of lose trust in it. And this is something kind of more oppressive for regimes will take advantage of and use to punish people who are showing genuine information. That's interesting. It kind of speaks back to what we were talking about with the, one of the Russian strategies is just throwing out as much information false and true as possible, right? the bigger danger is for people to become cynical.

Yeah, I mean, in a way there has to be a certain level of cynicism. It's been interesting in the last few days looking at Boris Johnson and this kind of argument that was reported with his partner. And then this photograph that was published, I think, in the Daily Mail, showing him and his partner in this lovely countryside scene, looking like they were getting along. And, you know, the male kind of spinning it in this very positive way for Boris Johnson. But then people were looking at that and, you know, even journalists were saying, well, it looks like his head's a bit longer than it was yesterday.

And was this actually a recent photograph? And when he was interviewed about it, he seemed very evasive about exactly when it was taken. So, you know, it might be in that case the cynicism is justified. But what we're trying to do with Balancat is kind of crowdsource the geolocation of that photograph to see if we can confirm wherever or not it is actually old or not, rather than just relying on cynicism to kind of, you know, lead the way. Well, I think that that's actually a hopeful note to end the show on. We don't normally do a lot of those. Elliot Higgins of Bellingcat, thank you so much for coming onto War College and walking us through all of this. Thank you.

That's all for this week, War College listeners. War College is me, Matthew Galt, Kevin O'Dell, and Derek Gannon. It was created by myself and Jason Fields. Special thanks to Samantha Cole, a colleague of mine at Motherboard, who gave me the excellent questions about deepfakes. As always, if you like the show, please rate us on iTunes and leave a comment. It helps other people find us. We're on Twitter at War underscore College, and we will be back next week. Stay safe.

Until then.