 I'm an open source investigator and trainer, okay? So yeah, being open source investigator, I do investigations. I come from Kyrgyzstan, so my main focus is in Central Asia, but I'm also looking into conspiracies in Europe and data-driven investigations and many other things, as well as the war reporting in Ukraine, but also in Tajikistan, okay? And yeah, let's dive into our presentation. This will be a little bit interactive at some point, okay? So I'll be needing sometimes for you to raise your hands. So yeah, the first question when we're talking about open source investigation is what is open source investigation, of course, what is open source research? And some of you might know, it is an investigation where a researcher and a reader have equal access to the information, okay? So how it happens in traditional journalism. Researchers might go do a field trip, might do an interview, or have some sources they can reach out to, right? But in open source investigation, researchers such as we, we're doing everything at our laptops and we're searching information that's available publicly. So theoretically, if I do an investigation, you would be able to do the same thing, okay, to reiterate each step. So this is the main difference. The researcher, which is a journalist, analyst or investigator, has the same access to information as the reader slash listener, okay? In our case, we include the information with a payroll being a open source. So this can be a subscription to a news channel. So say you want to subscribe to an economist or Wall Street Journal, it costs a few bucks. And of course, like in many cases, there is a high possibility that somebody else has always written about the subject you're interested in, yeah? And then when we want to check out this information, of course, we subscribe and we read this info. And we think, yeah, this is also an open source information because everyone with a budget would get access to it. So this can be as cheap as a couple bucks, but also expensive as being a couple thousand of dollars for a satellite imagery. So we use satellite imagery a lot at Balancad. It's primarily Googlers, but we also have paid subscription to services like Planet, which costs, yeah, I think about a thousand, $2,000 a month, but this gives us an access to a huge amount of satellite imagery, okay? So there is this, I'd say, budget barrier, of course, but we still consider that it's an open source. And why is this important? Because when Balancad just launched in 2014, so it's a relatively new organization, to build trust, among our readers, we, our founder, decided to do everything open source because he really liked it, he was really into it. And then he could see that when he did an open source investigations, readers could actually look at each step and confirm that, yeah, this is a trustworthy information. So the reader doesn't have to take our words for an investigation, they can just find the same photo or delicate the same photo and do the research themselves, okay? So yeah, the great thing about open source that it provides transparency and the credibility to this investigation. And yeah, in an ideal world, again, consumer, the reader will be able to do the same investigation. So open source materials, what are those? So we have these four blocks of open source materials categories, okay? So first of all is the GSPayShell, the offensive word for mapping services. So these are the things like Google Maps, Yandex Cartier, which is basically Yandex Maps, Terra Survey, Planet Subscription, and etc. So this one is very, very useful for many things. One is, of course, for geolocation. So if you have a photo and want to find out where exactly it was taken or you want to verify that this image is showing, I don't know, which is showing the Netherlands and not Belgium and things like that. So in this case, mapping services, they help a lot. Another thing that you can do is, for example, just tracking the changes. So for example, on Google Earth Pro, so you know, on Google Maps, it will show you the latest satellite imagery. But on Google Earth Pro, which is the app, you can see actually the historical images. And then you can trace the historical changes, basically. You can see if some construction has been made, if the mining project is going on, if some forest fire coat up and the consequences could be seen on the maps. And many, many different things. The second one is media. Of course, as I've said, there's a high chance that media has always republished something about the research topic you're investigating. And that's why we always check out what's there. So yeah, these are just all the different medias that are out there. In most of the cases, how our research works is that we are reading some article and we see there mentioning some photo or some interesting thing that they didn't follow up on. And we say, oh, that sounds interesting for us. We could probably do something about it. And then we base our research on it. The third one is user-generated content, which is basically everything on social media that users are posting. You all perfectly know how much information is out there on YouTube, on LinkedIn, on Instagram, TikTok, Facebook, Vkontakte, and et cetera, et cetera. So we use this information a lot. And the last one is databases and archived materials. So basically human knowledge of the world. This can be databases of books, like Google eBooks. It can be governmental databases of companies, like Companies House. It can be a private database of companies, like Lexus Nexus, which costs some money, information from Internet Archive, which has an archive of millions of websites, but also books, videos, and et cetera. So these are all the main four categories where we get our information from. And Joel quite used to those. And here are some other four categories that we do use from time to time. This are called gray area open source materials. Why gray area? Because there's a lot of question about ethics about them. Not all journalists consider these sources as open source. Some journalists do, some journalists don't because of the ethical reasons. So for example, we do use leaked materials. So this can be information from WikiLeaks, but it can be also information from leaked passwords and leaked platforms, like from D-hash, intellects, et cetera. It can be databases that we can buy basically on the Internet or find on Torrent. Here's an example of a database of Russian airlines. So Russian airlines, which is called Iaraflot, a couple years ago, we had a huge leak. So you could find info about the passengers. If you're searching for a person, you could see their past flights, their date of birth, their passport information, the people they flew with. So basically the people they bought tickets together with. So this kind of information. So if it has been leaked, we might use this material depending on the investigation. So I'll talk a bit about in the next slides when we do justify using those. Because there is, of course, two sides of it. So in one case, it can help us to find this GRU agent, FSB agent, or any other bad guy. But there is also another ethical issue is that because by buying this information we're incentivizing, of course, hackers to get this data, to leak this data, or we are supporting this platform which shares personal information of lots of people. So there's this trade-off. And in each case, we always have a discussion if it's really worth looking at this data. So we wouldn't just check it for some random person, but only when it would be in the public interest. Okay. Next one is contact books, which is apps like Get Contact, True Color, Sync Me. Has anybody heard about them? No? Okay. Some of you might, I think, decide the apps that you basically don't want it to your phone and you give them access to your contact book, okay? So when somebody else, so the great thing about it is if somebody else that you don't know is calling you, it will tell you the name of this person, okay? It can tell you right away that it's a scam, or it can tell you right away that it's a bank, or some person's name. So how this works is, as I said, when you download it, you give access to your contact book. And when there are lots of people who have downloaded it, the app will have a huge contact cloud, right? And then you can do searches, you can search for this person, and then find how this person has been saved in some other users, okay? So for example, let's say that I never downloaded this get contact book, right? So I never downloaded it, I don't have my information on it. But my sister, for example, she doesn't hear, okay? She doesn't care about her, like, contacts being leaked, or she might have not read the terms, and she agreed to that. So if she saved me as, like, sister, okay? Then if you would search for my phone number, you would find that someone has saved me as sister, okay? If she saved me as ironish, I don't know, like ironish, my name, then you would find that someone saved me as ironish, okay? So this is a great way to find information about people, because usually this, like, bad guys, or just in general, like, people who don't want their information to be public, they have no say on it, right? Because if my sister, if my friend, if my colleague downloaded this app and they had my contact, then now my information will be exposed there, okay? So, yeah, this is, this is a sketch of things about contact books, and that's why they're in this gray area open source. But we do use them quite often, again, when we have a phone number and we want to find out who this phone number belongs to. The next one is exposed databases, database on MySQL, Amazon Web Services, but also, like, Google Sheets and stuff, and these are the databases that users have, but they don't really know that they made this database public. And then, again, there's this ethical question that, are you gonna use this information when the user didn't actually know that they made it public, okay? And the last one is dark web markets, like, Silk Road, Dream Market. We don't often, we don't really do our research there, we just, this applies for us to sometimes download like databases, but yeah, that's pretty it. Okay, so what is open source? This is a source that is freely available to the public. As I said, this can include a payroll, but in most of the cases, we do use free tools, okay? And I'll show you, I'll show a few at the end that we use, okay? And this does not include directorying or requesting hacking, stealing or spying, but it can include accessing publicly available leak data, okay? So we're never gonna, like, come up to a hacker and say, hey, could you hack this person, social media, or hack this database? But what we're gonna do if some hacker has already leaked it, if some organization has already leaked it, we might use this information, okay? So yeah, and just remember that when we are doing a research, it, yeah, the results that we have will depend on lots of different factors. If you are looking into a person, it will, of course, will depend on how much digital footprint this person leaves. We will also look at the platforms this person is using. So for example, if you're looking to Vkontakte or Facebook, Vkontakte is basically a Russian Facebook. So if you're doing a research on Facebook, and if, you can't really go out of Facebook to many other places, whereas if you're researching on Vkontakte, even if the account was closed, you would still be able to see their friends list, you would still be able to find out their date of birth, their alt profiles, alt bios, and et cetera, just because Vkontakte is more research friendly and it can get scraped, okay? The same with WhatsApp and Telegram. So Telegram is also a messaging app similar to WhatsApp, but it just has more features, and again, it's more research friendly. And if you were doing an investigation based on WhatsApp, it would be extremely difficult because you can't really find groups, you can't really find different channels and stuff, but if you found a Telegram, you can, first of all, do a search on different groups, even on private groups, on channels, private channels, and stuff like that. And if you find an account, you'll also be able to find their user name, and then just continue and continue your search, okay? So it really depends on what platform they're using. Another thing to remember is the privacy legislation in the country they're residing in. So for example, same with Russia. So in Russia, the government was collecting lots of information about the citizens, and they were intending to do this private, and they keep it to themselves for their use, but it gets hacked all the time, and that's why you can find a person, like a Russian citizen's passport information easily, their flight record, their cars, where the car is registered, the phone number that they used, and etc. So lots of information there. So if you're looking on information about Kyrgyzstan, it's somewhere in between, you can find some information. So if you were to research me, you would find my full name, my paternal name, you would find out my date of birth, my relatives, the village I used to live in, Kyrgyzstan, and things like that, because this information was made intentionally public, okay? But you can't really go further than that. Whereas in Germany, for example, it's completely closed off, because their privacy and data legislation is much better than I'd say. So it really depends on the countries you're researching in. Okay, yeah, open source investigations called a new genre of journalism and beyond. Again, when our founder started Bellinghead, it was 2014, it was quite new, and he was one of those pioneers, and nowadays it is really caught up, I'd say. So for example, New York Times has a great visual investigations team, where they do great just investigations into different matters, and they've won a couple of pool surprises for it. Washington Post has also their visual forensics team. BBC Africa Eye has also done some great investigations. So here there was this investigation called Anatomy of a Killing, just an amazing investigation that BBC Africa have done. So what happened was that there was a Facebook video circulating on the internet with some soldiers executing women and children. So there was like this horrific video. It was allegedly shot in Cameroon, and the people, the executors were Cameroon soldiers, okay, allegedly. But the government would deny this, saying this is fake news. I don't know, this is from a movie or something or some other different place, so it's not true, they were basically claiming. And then what BBC Africa Eye did, first of all, they geolocated this place. So you could see just a tiny bit, just a short video showing the soldier walking with women and children, and you could see some place around, like the road, how the roads looked, what buildings there were, and et cetera. So they were able to geolocate it and confirm that yes, this was actually shot in Cameroon. And then they were able also to confirm that this were actually Cameroon soldiers. And you guess how to do that? Okay, I guess, yeah? Yes, so looking at their uniform actually was very easy to check. What they did was they just googled like Cameroon uniform and they could see that this match is exactly the same. So they had to be on the ones. And yeah, as the result of this investigation, it was confirmed that this video is true and the children and women were executed by Cameroon soldiers, and later they were convicted of murder, okay? So this is what verification and geolocation can lead to. This is also getting popular in the non-governmental organizations, in the human rights organizations. So here's, for example, a vacancy for a human rights watch for open source investigations had. UN also has a team of open source investigators, okay? Again, this is all because people know how much power this open source has hold. You don't have to necessarily be in the conflict zone or be on-site to do investigations. You can sometimes do it just online, okay? Searching for relevant videos on the Internet, searching for relevant videos in the local language on YouTube, on TikTok, and et cetera. And yeah, and since there are so many information that is out there on the Internet, you can apply it to a great range of topics, okay? So this can be conflict zone, and conflict zone usually are the ones that get the most attention. They're not, I wouldn't say in Israel that this are the most popular investigations that are there, but that researchers do, but this are the ones that get the most attention. So for example, right now at Balanket, we have this project where we're looking at the Russian invasion of Ukraine, and what we are doing is we are basically collecting all those photos, videos posted in Ukraine, okay? So yeah, as you know, there are lots of people who are just taking photos of damages, of murder, of different atrocities, and we at Balanket, we are collecting those. We are verifying that this indeed happened in Ukraine and also it is new, it's not from like 10 years ago or from another place. We are also geolocating it, confirm that yes, this was, like this video was done in Lysichansk, this hospital was located in Lysichansk, this number like no five schools, et cetera, in Kiev, and et cetera, okay? So we are trying to geolocate it, and then we have this nice database of all these incidents. We are specifically focusing on civilian harm done in Ukraine. And then this gets transferred into a nice map where people like anybody can access it and see the cases, the incidents of civilian harm that we were able to verify and geolocate and confirm, okay? And why is this useful? It is useful because the, first of all, like other researchers, other journalists can do their investigations based on this data. The government, different justice accountability bodies can use this information because this has been verified and vetted by our staff, okay? And lots of other places. So for example, I'm currently looking also at the protests that turned out in Tajikistan, and yeah, my colleagues have looked at Yemen, Syria, and also many different other places, okay? So this can be applied to any conference. Yep, the next one is environmental research. Again, quite popular, I'd say. Just again, using different mapping services. You can see the changes in, I don't know, in the sea level or the level of the lakes. You can see the frustration, many different stuff. But you can also look at the mapping services not only with a satellite imagery, but also with different radars to see how the moisture index has changed, how the vegetation index has changed, and et cetera, okay? And lots of other different things. Basically again, looking for photos, for videos, posted on YouTube, Facebook, and et cetera. Then there's corruption and financial research. I've done a few investigations to corruption in Central Asia. And this are the cases, for example, in my case it was the fact that there was this very, very well-dandoned, influential corrupt family, but no one could prove that they were corrupt, right? And what we were able to do is find their Instagram accounts and Adna Klassniki accounts, which is like basically, again, Russian Facebook, of their family members. And we could see that they have, they're having this luxurious lifestyle that they would not be able to afford on their government salary. We were also able to find their nanny of this particular family, who was posting photos with private jets with the kids of this family who was doing a whole video tour of their new apartment in Dubai. Hold ten minutes of saying, like, my bosses have just bought this apartment, and we could see from their wife's Instagram account that this is her apartment because she posted videos. The walls were the same, the doors were the same, and different furniture. And then we could confirm that, like, when this nanny posted a photo from here, from this penthouse window, we could gel-locate it, we could find it dislocated in the most expensive area of Dubai, in this building, and it was a penthouse, and this cost a few million dollars. And again, this family was making around $10,000 a year on their governmental job, being the head of taxes in Kyrgyzstan, and there was no way they were able to afford it. So this came as a crucial evidence, yeah, crucial evidence to actually prove that they are corrupt, and then later they had to confirm it, and they, yeah, he was basically arrested, okay? There are also things like HR, criminal, a person-focused research, of course, and historical research. And the fun fact is, like, brands actually the first open source investigators because they used the information that was publicly available. So, yeah, and lots of also other very different things, okay? It's just the ones that are more popular right now. So, yeah, and what is Bellingcat? How many of you have heard about Bellingcat before? Okay, okay, I'll skip this one. So, yeah, we have our website. We publish in English, Russian, and Ukrainian. Some articles get translated. Some of the most popular ones are our investigators in Syria. So, Elad Higgins, our founder, he started actually with Syria. In his free time, he would just research what's going on there. It was like during 2013, 2014, because it was very difficult to find information about the Syrian war because it was... For journalists, it was very risky to go there. And what he figured that he can actually, like, find videos of what's going on there in Arabic because local people were also posting photos and videos of what's happening. And journalists just didn't know how to find it. And he was just, yeah, basically searching the same terms in Arabic. He would able... He was able to find videos. He would gelocate them and confirm that, yeah, here are chemical weapons being used and stuff like that. So, he had actually no background in anything like that. He never had a... Yeah, he never knew how to differentiate among different weapons, but then he just, by Googling, he became an expert at the end, okay? There is MH-17, the Downing of the Malaysian Boy and Cobra Ukraine in 2014. Again, when this one happened, no one really knew what was going on. Russia was blaming Ukraine. Ukraine was blaming Russia. And no one knew who was responsible for killing almost 300 people. And what we decided to do at that point is again turn to social media. So, you all know that the boy cannot be shot by just a random weapon. It has to be a missile launcher, right? Rocket system, some. And there are not that many missile launchers in Ukraine and Russia. And we suspected that if there was a missile launcher traveling, it would be with a convoy, with a military convoy. And some people would, of course, see it. Some, like, citizens would be able to see it and then upload it to the Internet. So, we started just searching on YouTube, on local social medias for a keyword military convoy. And then we started finding photos and videos that people were actually seeing it a few days, a couple of weeks before the downing of the boy. And then we were able to delicate those and then make this map. So, this missile launcher was spotted in Russia going to the border at this point of time. Then a couple days later, it was spotted in Ukraine at this place at this point of time. And then we could see the textual travel from Russia. And after the downing, we could again find the photos and videos of this missile launcher of the book. And when it was going back, we could see it was missing one missile, okay? And, yeah, this is just an example of how we looked at it. We also looked at the military unions that were near Ukraine, okay? So, one of the closest would be Kursk, which is in Russia. And then we started searching for soldiers who worked at this Kursk military unit. Because same as on Facebook, like people, like these people are like 19, 20, 21. They would, of course, indicate in their bios on their social medias that they work at this military unit. And then we would find their social medias. We would find that they were actually taking photos with their book. They were in this convoy posting photos because for them, it was quite exciting. It was probably, again, they're like in their 19, their 20, they must have been their first trip. So they have uploaded lots of pictures on social media. And it was 2014. They had no idea that this info could be used, okay? So, yeah, then we have Scripple and Navalny Poisonings, where we've actually used quite a few of this gray area sources. So there was poisoning of Scripples and Navalny by Nevychok. And, yeah, again, in most of the cases, it was not really clear what happened. So what we did, we used this gray open source area, so leaked databases. If we had a name, we would search for this name and see if this name has any car registered to it, passport information, et cetera. We would find different flight records of these people. So for the, I think it was for Navalny investigation, we were searching for people who were traveling with him on the same route. And we eventually found some people that turned out to be traveling with him for a few years so they were tracing him. And then, yeah, we could find their passport info, their full name, their phone number. And then we'd go to those contact books, like Get Contact, True Caller, and searching for this phone number. We could see that. So this is a telegram bot, okay, which basically works on Get Contact. And we uploaded a phone number and we would tell us how this phone number was saved. Again, this was a phone number of agents, probably, and then he for sure did not have this app downloaded. But some of his friends did, some of his family members did, and they saved him as Vladimir Panayev, Panayev Vladimir, and FSB, Vladimir Alexandrovich Panayev, okay. So these things happen a lot, okay, saying like this FSB, because people usually save them as like where they know them from, right. And then we look at their phone records with this Navalny investigation, since we had a phone number, we went a little step further. So in Russia, you might know there's a lot of corruption going on same in Kyrgyzstan and in most of these countries you can actually find different websites of the workers from mobile operators, okay. They work for those mobile operators and they'll get you the phone record of a phone number that you're looking for, okay. So we requested this information that did cost us some money, but we could find out like all the phone numbers, this phone number called. And of course it would be like other high-ranking FSB officials, there would be a commander and stuff like that, but we also could see their location, okay, where this was, where the person was when doing a call when the phone was on. And yeah, you should remember that this is like FSB agents. There are also people, they do mistakes, some dumb of them. So for example, they have this rule, they are not supposed to turn on their phone wearing a mission because they can get tracked. And what they did actually was when they were near the hotel of Navalny, they actually turned on their phone for three seconds and these three seconds were enough to get the location from the mobile operator. And since we bought this data, we could also find it, okay. So let's do a quick example, which is called Locating ICS supporters. 2016, you might remember there was a huge campaign by ICS supporters taking photos and saying that they are actually, they live in Europe, they're in this city and they're actually among, yeah, and they're not basically, they're not far away, but in the Munster for example, in Paris to spread terror, okay. So what you would find is this kind of photo posted on Twitter and this one says like ICS supporter, like Germany, Munster, okay. And lots of ICS supporters would do that to spread fear that again, this ICS supporters are not somewhere far away, but here in their cities, okay. And the open source community started gelicating it and let's do it together. So please, yeah, just shout out a place that, shout out the details you see that can help you find out where exactly this is. Yeah, with this Munster, yes. Oh yeah, sure. So we want to find out where exactly this photo was taken. So what are the details that will help you? We have a first test for saying the bus. Yeah, you can actually just Google Munster bus and find out that Munster has similar buses. We cannot see really the number, okay, because it was uploaded on Twitter, so the route we didn't fill now, yep. Yeah, there is a poster and this is also very important. We couldn't really find, figure out what was written there and what advertisement it was, but then it got very useful, okay. Yeah, yes. Yeah, so now we're looking at the micro details. So if you find some similar place, the details you would cross reference, okay, to make sure this is the right place. So yeah, there's the white building behind, what else? Uh-huh, you can see the trees, you can see the lights I've heard. Yeah, you can see that there is one light there, two of them, yes, there is also a big shadow, okay. What do you think it is? It might be a bridge, it might be a building, but there, we'll go back to it later, but if you look at this place, it will give you a hint what this is, okay. So yeah, we could see also the crosswalk, okay, the crosswalk here, a few trees here, and yeah, these are the things. So we couldn't figure out where this place is, so what we did, we posted on Twitter saying, hey, does anybody recognize this place? Does anybody know how to find this place? And then this German resident tells us, on Twitter just replies, saying, actually in Germany, there is a map with all advertisement stands, okay, and then you can find it. So we went to this website, you can see there are just hundreds of these advertisements columns. So we go to Google Maps to do a satellite imagery, and then we start filtering the ones that could be, okay. Now, I'm gonna give you six options, okay, and you'll tell me which one is right, okay. Now look again at the details in this photo. What do those lights tell you? Yes, there is a cross, yeah, there is an intersection in there, okay. So look at all the details, okay, let's go. The first one, do you think it is the right place? Why not? Yes, no crossing. By the way, this green thing means this is where the ad was, okay, the column. What about this one? Okay, I'm hearing a no, let's look in the third one. Oh, it might be two lights for you, yeah. I'm sorry? Yes, there's no white building across the street, that's right, okay. What about the fourth one? Oh, I'm sorry, not anyone can see. Oh, perfect, that makes sense, yeah. What do you think about this one? Okay, I'm hearing maybes, okay. Number five, what's that? Yeah, yeah, it's not roundabout because they don't have lights, that's good. And the last one, okay, who thinks it's the number one? Number two, okay, I've got one, maybe. Number three, number four, okay, let's look at number four then. So now we are gonna zoom in, okay. And now we are gonna look at the micro details, okay, on the small details that match. So what do you see matching? Yeah, I can hear traffic lights. So here's one traffic light, yeah, and then here's the another one, and you can see it's also this shape, okay. This, yeah, so here, here's the shape. You can see it's not straight, it's curved, and here's the same. Anything else? I'm sorry? Four legs? Oh, that might be tricky because this is satellite imagery from Google and it was taken I think a few years before this took place, okay. So yeah, unfortunately, maybe you can do it, but we cannot access the satellites that are doing it the exact time, okay. So, yeah, yes, the lines. So you can see there's one sick line here, the same as here. I'm not sure if you're able to see it here, but there was also another one dashed, the thinner one. And you can see it's here too. Yep. You can see also this barrier. You might think it's a shadow, it's actually a barrier. And then we have a question about the shadow. Does anyone have a guess what the shadow means? This huge shadow, huh? Direction of what? Yeah, yeah, why is it so huge? Yes, that's right. So this photo was uploaded around 7 a.m., 8 a.m., and the shadow is much longer than it's in this Google photo. So the satellite images are usually taken at noon, sometime close to noon, to get the least amount of shadow, okay. So here's just an example that you can also track different projects, okay. And yeah, now we're gonna do another exercise, but just give me a second. We are finishing at 12, right? Oh, okay, so I thought 12. Okay, let me give you another example about the overflowing of information. So open source investigations are not black magic, okay, so we can only find what's there out in the Internet. But it also has its own challenges. So one challenge is that you cannot find that information that you need, meaning that there's just so little information about this person there or about this friend. Another is the opposite, having just this huge amount of information, okay. So here's an example. 2017, this white supremacist protests in the Charlottesville in the U.S. A photo gets posted saying that three, this three white dudes beat up a black kid and then this black kid ended up in a hospital with lots of injuries. And we wanted to find out who this person is, okay. So all we have is this one blurry picture and we know that this is a Charlottesville event in August 2017. So how do you think we can find this person? What would be your first step? Yeah, I can see the helmet here and there. So you can see this, he's wearing a helmet, white helmet. If you look closer, you'll also see that he has stickers, black and red that makes it quite unique, okay. That's good, so you know how to differentiate him from other people, yeah. Yeah, exactly, that's exactly what we did. We started looking for other photos and videos published about this event. We did like hashtag Charlottesville on Twitter, on Facebook, but we also just Googled and tried to find out as many photos and videos as possible, okay. So eventually, Eric was able to hear, yeah, he looked through a thousand of images and he eventually was able to find this guy. Do you think it's the same person? Could be. Why is that? Yeah, shirt is the same I'm hearing. Yeah, stickers also match, I'm hearing, yeah. So the shirt matches and the stickers also match, okay. So you can see there's this black sticker, but also red sticker. And also if you zoom in, you'll be able to see that there's something written in here and yeah, and this one's also match saying something killer, maybe. Yeah, something like this. And yeah, but again, so now we found a couple of images of him where we can see his face. But again, on this images, there was no text, no user names, no names, and so forth. So what we do is, yeah, we just continue searching for this person and then eventually we find this guy. So I'm pretty sure that some of you would suggest doing a facial recognition, but this one was back in 2017 and facial recognition in this case did not help. But nowadays, I'm pretty sure you would be able to find something. So then we find this guy. Do you think it's the same person? Yes, because of the birthmarks, that's right. So yeah, the birthmarks are the same and he's also wearing a chain and yeah, if you look at his ear, it's also has the same form. But again, we found another picture of him, but we didn't know who he is. So the next step for us would be searching for his friends, okay? It was the exact same, the exact same thing because we could see in the photos that he was always with this guy. So we assume that this are his friends and we find, if we find information on them, then we might find information on them. So we do the same thing and then we find this Twitter account called YesYourRacist basically exposing all these white supremacists on the protest, okay? So they posted this photo and screenshots from Facebook accounts of these guys. So we go to their profiles, we find Jacob Dix. You can see that the same guy. We can see Ryan Martin. Do you think these are the same people? Why is that? Yes, that is, okay? So here we're 100% sure that this is the same guy. We look at their friend list and then we can see that Jacob Dix, the blonde guy, his friend was Ryan Martin and then we just go through his friend list because we expect that the first person to be his friend and we find this Dan Borg-Borden. Here is the guy. We can see that his friend was Jacob Dix and Ryan Martin and then here are photos of him, okay? Do you think it's the same guy? Yeah, Burzmarks. So you can see that he has the same Burzmarks. He's wearing the same chain. And yeah, if you again look at the ears, it will be the same. Yeah. Yeah, and this is how just by sitting at computer area, I think it took him around two weeks maybe. He was able to go from this blurry photo to actually the name of this guy, okay? And then about two months after the investigation, the guy, Daniel Borg-Borden, was arrested and found guilty, okay? Yeah, so other challenges is that ethical concerns, okay, that I've told you about. Where do you draw the line on gathering personal information on someone? So, for example, if you're looking... If you're looking into neo-Nazi groups, do you think it's okay to create a fake profile with some other men's photos and pretend they're also watching the premise to get into this group and then infiltrate and check out all the data, okay? You might find it's okay. You might find it's not okay. Or if you're looking at this person, are you going to create a fake profile of their friends to friend them? And then once you get access to that information, delete it immediately. Again, I think most of the people would say no, but still some journalists would do it, okay? So it really depends on the line where you draw... Where we draw as a researcher, as an organization, and, of course, depending on the law of the country, okay? And yeah, if you're interested in this open source investigations, you can look at our tools. We have our toolkits at this link, because, by the way, I'm going to send you the... I'm going to send the percentage to organizers so you can get it, okay? And also, there is a great book of Michael Bezos, open source intelligence techniques. It's very thick, it basically has everything... Yeah, everything that a researcher would need. So yeah, sorry, we're a little bit late for questions. Yeah, thanks. I just checked the program, and I think the next one up is at one. So I think we have some time for questions. Yeah, let's just use it. Yeah, so... Okay, thank you for your presentation. I wonder how many languages do you master, and how do you approach languages that you don't speak or can read? Mm-hmm. I'd say some of us speak only one language, English. I personally speak three languages, and it really depends. What we use is Google Translate, okay, basically. So the team that works on Ukrainian research, like, I think most of the team, of course they don't speak Ukrainian or Russian, we have a new Ukrainian researcher, but still, majority of us are just translating it from Google Translate, yeah. So that's the thing. So first of all, thank you for your work, giving us some proper journalism to read. So two quick questions. First of all, how do you deal with, I guess, individual and organizational security and privacy? And secondly, how do you deal with organizations, individuals, intelligence, whatever, that are basically pretending to be an adversary, like planting evidence, creating videos, uploading them that are pretty, like, damn realistic? And how do you, like... Yeah, sure, good question. For our security, yeah, we take it very seriously because we always get phishing attacks from, yeah, different people, especially from the Russian government. And for that one, we have our, first of all, we have our security person and we follow the protocol. So one of the thing is having the perfect cyber, having our digital security very, very well maintained, not clicking the links, not having the same password, so the very basic stuff. But also for the physical security, I'd say there's only one not visiting Russia. For, like, smearing campaigns, sometimes it does happen. In this case, we do have a trauma risk management policy where if a person is getting a rest or something like that, they, first of all, can approach to a psychologist, but we also have editors in the, yeah, and basically the management staff who can make a statement or who can take care of it or who can decide that it's better not to reply to the Twitter user, for example. That's a very question. Oh, okay. I'm sorry? Gami, one question is enough. The other two you can ask later after. Yeah, yeah, I'll be here. Okay, perfect. Because this will be the final question because there are people standing on their legs for an hour already and I think we really need to wrap this up if that's okay. Go ahead. Hi, I was wondering if you're estimating the uniqueness of, let's say, the stickers on the helmet and the birthmarks because there might be some collisions at some point. Say it again, sorry. There might be some collisions with other people having similar birthmarks or similar stickers. Do you estimate that? Yeah, yeah. So what we can do, so first of all, we look very closely at the birthmarks. There are actually hit around 10 and they were perfectly aligned. So in this case, we're 100% sure that the same guy, but if a person doesn't really have this birthmark or a tattoo that would give us 100% of that same person, we would go to, for example, like Microsoft Azure's tools to compare faces. Okay, so yeah, tools basically to compare faces. If it's higher than the 70% match, then we say it's the same person. So this basically, yeah, all good. Okay, thank you very much. I do have a workshop on geolocation in one hour and then you can also right now catch me for other questions, okay? Ooh, that's cool. Yes.