In deze aflevering
Eliot Higgins was in 2011 werkloos en vocht, zoals hij het zelf zegt, vooral ruzies uit in de reactiesectie van het Midden-Oosten-liveblog van The Guardian. Tot iemand vroeg: hoe weet je eigenlijk waar die video is gefilmd? Higgins zocht de moskee uit het beeld op in Google Maps, won de discussie en ontdekte een methode. Als blogger Brown Moses groeide hij uit tot pionier van open source-onderzoek. In juli 2014, drie dagen voordat MH17 werd neergehaald, richtte hij Bellingcat op.
In dit Engelstalige gesprek vertelt Higgins hoe een handvol vrijwilligers uitgroeide tot een onderzoeksorganisatie met bijna veertig mensen wereldwijd, sinds 2018 als non-profit geregistreerd in Nederland. Hij neemt ons mee door de grote dossiers als MH17 en de vergiftiging van Skripal en Navalny, maar ook door de absurde verhalen eromheen: Bulgaarse spionnen die hun eigen inbraken minutieus documenteerden in een Signal-groep, en een gestolen hond die dankzij Bellingcats kentekenanalyse binnen een kwartier terecht was.
Voor mediaprofessionals zit de kern in de tweede helft. Higgins betoogt dat redacties niet langer bovenaan een informatiepiramide staan, maar onderdeel zijn van een netwerk, en dat ze hun publiek moeten leren zien als deelnemer in plaats van ontvanger. Transparantie over je methode is daarbij geen bijzaak maar de bron van vertrouwen: iedereen kan het onderzoek van Bellingcat zelf narekenen. Over AI is hij ambivalent: nuttig als gereedschap voor geolocatie en “vibe coding”, maar het grootste risico is dat het mensen een excuus geeft om échte informatie niet meer te geloven.
Deze aflevering is Engelstalig.
“It wasn't like some noble intention. It was just fighting with people on the internet.”
Eliot Higgins · 6:15
Hoofdstukken
- 0:00 Introductie: wie is Eliot Higgins?
- 3:02 2011: een ruzie op internet als begin van een methode
- 8:29 De oprichting van Bellingcat, drie dagen voor MH17
- 14:30 MH17 en het Joint Investigation Team
- 19:54 Kickstarter, Hugh Grant en de Postcodeloterij
- 24:44 Navalny, Skripal en de Bulgaarse spionnen
- 35:13 Data van de Russische zwarte markt
- 41:01 Van gatekeeper naar netwerk: een nieuw informatie-ecosysteem
- 52:09 Transparantie bouwt vertrouwen: meteen publiceren
- 56:04 Een haat-liefdeverhouding met AI
- 1:00:25 Wat redacties nu zouden moeten doen
Transcript Lees het volledige transcript
Dit transcript is automatisch gegenereerd en licht geredigeerd; er kunnen onnauwkeurigheden in staan. Tijdstempels linken naar het fragment op Spotify.
Okay, Philippe, we have an international guest again. That's why we are speaking in English again. And, it's somebody whom we have imported from the UK. He's a guest from the UK. Well, I think he's very interesting in what he did on a journalistic level, but maybe you can introduce to him.
Yeah. It's a great honour for us to receive today in our podcast, the legendary Eliot Higgins. He is the inventor or at least pioneer of open source investigation in the internet. And that's using social media, flight data, satellite imagery, all kinds of open data that you can access through the internet for investigations, for journalistic investigation. And, people, know him as the founder of Bellingcat, a collective of investigators that do this kind of work. And they are very famous for actually proving, that a Russian army shot the rocket that, that got down the MH17 plane, but also the poisoning of Skripal in London, of Navalny in Omsk, Ukraine war. There's all kinds of famous dossiers where Bellingcat, played a large role. But if you see the website today, it's full of investigation that you would never have dreamt of.
So this already started when you were like the editor in chief of de Volkskrant.
Oh, yes. Yes. Back in those days, in the teens, Bellingcat was a household name for a whole new kind of journalism. And people also in my newsroom started to work in this way. So it's been very influential change in journalism. And, if you look at the website today, you see it ranges from subject as the use of generative AI by the governing party of India to foment hate against Muslims, to telling explosions from each other. Because we have a lot of videos within our own war showing mushroom clouds and things like that, and things are said about that, that are inaccurate. And, if you, if you have to find the most nerdy specialists in this, knowing all the weapons and explosions from each other and able to tell it to us. He's working at Bellingcat and, but also tracing extremist rights, rightist activity in Europe, among them Netherlands and, of our understanding of how to research Chinese internet platforms. Very difficult. You're also educating, newsrooms, Bellingcat, and NGOs. So I'm very curious, how does Higgins see the role of journalism evolving in this area. What is AI changing in this and what can we learn from him?
Shall we get him into the room? Philippe?
Yes.
Let's get him here. Let's get him in here. Eliot Higgins, welcome. Welcome to the show. Thank you for being here. You are the founder of Bellingcat. As Philippe just nicely introduced us. I was always fascinated by, the rise of Bellingcat, but also the rise of the figure Brown Moses. You were the blogger Brown Moses and I read the book we are Bellingcat. And you wrote in there and also in a documentary, basically that around 2011 you were unemployed or in between jobs, as people sometimes say it. And you were sitting in your upstairs room, maybe in the attic, I don't know. And you thought, let's just go on to the internet and see what kind of data there is to, I don't know, investigate stuff. And from there, you started blogging because you discovered something in Syria or you discovered something, I don't know, something that was in the news. Tell me a little bit about this process, because this is the fascinating part of me. You are just Eliot, right? And then back then you were just Brown Moses. You were not a journalist. You were not educated like that, but just you just started like that.
Yeah. I can't say I had a very noble start or noble intentions when I started doing this. So I was in 2011 spending a lot of time, kind of looking at what was happening in Libya on forums where people were discussing it. One in particular was the Guardian's Middle East live blog, which had a comment section. It was probably only about 10 or 12 people who posted on it every morning, but it was just something I was really interested in at the time, and I would post stuff in the comment sections.
For no particular reason.
Well, I think as someone who grew up kind of with their, I think, childhood bookended by the first war in Iraq and then 9/11, there was always a kind of interest in international politics and how things were unfolding in the world. The Arab Spring was a huge event. But I think for me, what was interesting is, as someone who was probably spending too much time on the internet in general, you would see stuff coming from places like Egypt in 2010, like Tahrir Square, where there would be like a live stream of the protesters and then the police attacking the protesters. And we were kind of all watching it thousands of miles away, chatting about it on our internet forums. So for me, it was the first time that it was kind of like the first live streamed conflict, in a sense. And then in 2011, with Libya, video started being shared on places like YouTube that were from all over the country, but they were broadly being ignored by the media who didn't really know what to do with them because they'd been burnt before by trusting information on the internet, and they ended up being used for arguments, basically in comment sections, which was where I was at, using them in arguments and comment sections. But what was frustrating is I wanted to know what was actually going on, not someone's opinions of what was going on. And these videos were often being used as just evidence to support one side or another without any real analysis being done.
It wasn't that the question wasn't whether they were true or not. The question was whether it looked like it supported the position of the person who wanted to use it as a stick to hit the other person over the head with. So there was one video that was shared, and someone in the comment section said, well, how do you know where this was filmed? And I thought, that's a reasonable question. So I thought, well, there's a mosque in it. It's next to this road. It's supposedly in this town, so maybe I can see that on satellite imagery. So I went to Google Maps, looked it up, found a mosque with the same dome and minaret next to the same road. And I could tell it's kind of like spot the difference between the satellite imagery and the video. And I took that and went to the comment section was like, yeah, there you go. That's where it was. Ah, I win the argument. So it wasn't like some noble intention. It was just fighting with people on the internet. But I kind of then thought, well, this is actually good information. I actually confirmed something. So another video would come up and I do that again. And the Guardian live blog, kind of managers started noticing those comments and posted them in the live blog itself because they kind of built a sense of trust in what I was doing, even though I was just some random commenter on the blog.
And your name was already Brown Moses.
Yeah, it was a name I'd been using for years. It was named after a Frank Zappa song I was listening to when I registered to an internet forum, and then I just kept reusing that name on Twitter and on the Guardian Middle East live blog. And there's something awful forums. And, yeah, then I started a blog because I'd been posting stuff on Twitter. I found and I thought I need a bit more room to work with. So that's when I started the Brown Moses blog. And there I kind of set myself a task every day to post something every day. Yeah, but I'm not saying these were good quality posts. It was kind of like I found this interesting video from Syria, but because I was doing that every day, I started noticing patterns around the use of YouTube, Facebook and Twitter by media centers, armed groups and other groups in Syria. It was quite consistent. And I collected every YouTube channel being used by the Syrian opposition, which numbered about a thousand at the end of it. But it meant then every single day, I could see every single video being posted on these channels and look for something new and unique. And that's.
So what led you, after that to found Bellingcat? What was the reason you started that?
So I was blogging for a couple of years, and over that period, what I was doing went from my kind of weird hobby to something where there was a growing group of people interested in it.
How would you how would you define yourself in that moment? Were you like a blogger, a semi journalist, journalist?
Oh, I would never call myself a journalist. Even now, I would struggle to apply that term to me. But I was definitely very much a blogger. But I was doing something that was unusual because I was using that process. I was kind of developing myself because I was kind of partly writing for myself, not for other people. So I just wanted more information about the thing I was interested in. And this was a way to get that. But then I had people from, you know, news organizations, human rights organizations reaching out about specific posts that may have mentioned a specific human rights violation or a type of weapon they were trying to track. And I kind of built a network around that. I was invited into a Facebook group that was about Id'ing munitions. So then I started talking to them about learning how to ID munitions.
But were they asking you questions like, can you help us with the investigation.
A lot early on. A lot of it was, where do you find this video? Yeah, because they hadn't seen this stuff before. And one early example of that is in 2012, around the summer, I found a video that showed, cluster munitions that had failed to deploy and were kind of in a crater. And it was the first video that showed the use of cluster munitions in the conflict. And groups like Human Rights Watch and Amnesty have had a lot of concerns about their use. So someone from Human Rights Watch reached out and said, hey, can you tell me more about this video you found? And I gave them the information and then they said, if you, you know, if you see any more videos like this, let us know. And they wrote something up on their website about it. They didn't mention me. So then another journalist went to the person who wrote it and told them off about that because they knew I was the source. And then they started citing me a bit more in their work. But, for me, because I didn't speak Arabic, looking at the weapons being used was something I could kind of figure out using online resources, talking to these people. And that became then very relevant in 2013 when there was chemical weapon attacks in Syria, because I was the only person who'd spent the last year looking at every possible video with every possible weapon. And I'd seen some of these weapons being used in the past in other alleged attacks that were kind of didn't get into the kind of mainstream news around Syria. And because of that, I just built this, you know, broader and broader recognition around particularly, I would say, civil society actors and journalists working on Syria, because I was the only person really producing this kind of information from those kind of sources. But for me, I wanted a place where people could learn how to do this stuff, and also a place where people could come and submit their own investigations. And that was the idea behind Bellingcat. So we launched that in July 2014 with that goal.
Is there an explanation, according to you, why you were the only person to do this? Because normally when, when, when a new development happens, it happens simultaneously on several spots on earth, right? Without knowing from each other. But how would you look at that?
Well, there were a few different kind of communities that formed around this. So there was a Dublin based media and media organization called Storyful, and they had been set up. A few years earlier to basically license out viral content. So someone's cat video would go viral. So their Storyful would go to them and say, we can license this to news organizations so you can make some money out of it. And, with the Arab Spring, they had started looking at these videos coming from places like Libya, Egypt and Syria. And, they, in that kind of very small community of people who kind of came together online in, you know, 2010, 2011, they were kind of part of that. You also had Christoph Koettl from Amnesty International, who was one of the first people who understood the value of this. So we form this very small community to begin with. And we kind of shared ideas like the word geolocation. I'm still not sure exactly where that word originally came from in our community, but a lot like OSINT itself is a term I don't like to use anymore. Open source intelligence. But that was kind of just used by someone because we didn't have a word for what we were doing.
You mean Osint?
Yeah. Why don't you like to use that anymore?
Because it suggests the end product is an intelligence product, and it also is something that's used by basically posers on the internet to say, I'm doing intelligence. It's open source investigation or online open source investigation. I prefer not to use open source intelligence because it has that specific connotation to it. And open source investigation can be part of open source intelligence. But you know, those are two separate things in my mind.
Is it also because it might make you vulnerable, because people would perceive you are working for the intelligence agencies?
Well, there's two sides to that. Yeah, it could make us vulnerable, but other people think it's cool that we were doing intelligence and I o.
But for instance, the Russians were attacking you very much, of course, in the course of these years. And they also accuse you of, of, of, of receiving information from Western intelligence agencies. So.
Well, this is the thing I think the Western intelligence agencies would be very pleased if they could have been providing that kind of information at that period, because from everything I heard later on that I spoke to someone at an event in about, I think it was 2014 or 2015. And they were kind of from the private kind of intelligence industry. So, you know, commercial intelligence. And they said they knew some of the people in the traditional intelligence services just from their career. And I was their dirty secret because the kind of analysts would steal the content from my blog and then kind of repackage it and send it to their bosses. But their boss would have already read it on my blog because my blog was their dirty secret. They were e-mailing around to.
Each other, and actually, our Dutch intelligence service did a lot of work on Eastern Ukraine with MH17. But the court case was partly based on the information from Bellingcat.
It was very interesting with that because in I think it was October 2014, we had published several, several articles, I should say, when I say we in 2014 with Bellingcat, it's me and a few volunteers to be clear.
And you published it about MH17.
Yes. So they, the kind of team that had been formed around the investigation invited me to be, be interviewed at a police station, which is always a fun thing to be asked to do. So, I sat down with them for several hours and went through every single post we had published kind of line by line, like piece of evidence, by piece of evidence, explaining the whole process in detail in what we did. And they were kind of very quiet and just asked polite questions. And then I believe it was in January 2015, from what I was told, they then set up their own unit to open source investigation. So I think we were very influential in that moment showing that, yeah, actually, there's a lot of stuff here. You can have a process behind that. And that may have actually influenced the investigation team to set up that own their own open source team. And I think this was a very unusual case because there was so much open source evidence to look through with Syria. There was a kind of a very different social media landscape. You had limited access to. The internet stuff was generally uploaded to media centers or groups. You had access to, say, a way to smuggle the information out or a satellite phone. So you knew where to look to find stuff. With eastern Ukraine, it still had pretty free internet. You couldn't post stuff that was critical of the occupying forces, but you could discuss stuff you were seeing in your local social media group. So there was a massive amount of stuff out there that could be discovered. And that's why we spent so many years investigating MH17, because of the mass of data that was available.
And in the Netherlands, we had the joint investigation team. Did they also, support their evidence based on your findings. Would you say or did they try to verify what you did and what happened?
Well, the good thing about open source evidence is that you can independently verify it's just open. They could see the work we did. They could see the sources, and then they can go through their own process of verification. So one example is that we early on looked at the markings that were on the side of the vehicle, the missile launcher that shot down MH17 and showed that it matched a vehicle that was seen in a convoy in Russia a few weeks earlier. And we kind of drew our famous coloured boxes around the markings and stuff we normally do. When the Joint Investigation Team presented their version of that, they'd used this very clever grid system. They overlaid in like a 3D space in the images so it could be moved, and you could see it visually how they matched. And we were like, oh, that's a very clever way of doing it. But it was basically an analysis of the same pieces of evidence, just using a slightly different and I think better technique to show the match between the vehicles.
Okay. And the circle was not in their evidence. Your circle that they used, they didn't you use your pictures.
We did see though that they did release a video one time and they had one, the investigators with a folder in front of him. And it was printouts from the Bellingcat website that were on the pages. So we kind of saw that and thought that might be a little kind of shout out to us, snuck into the video.
So you started Bellingcat in 2014? Yeah, if I'm correct. Was it before or after MH17?
It was three days before MH17.
Three days.
Before. And believe me, that has launched a number of conspiracy theories.
Yeah, yeah, I guess so. Yeah, yeah. Because it also.
Shot you to fame. Yeah.
Put you on the map.
I, I mean, I think MH17 I'm often asked what are the most important investigations we've done? And MH17 is one because, first of all, it obviously brought a lot of recognition to the work of Bellingcat, but it also was a huge catalyst for the community around open source investigation that it brought people together because there were lots of people on the internet looking for information about what happened. And Bellingcat became a kind of, you know, a focal point for that work. Journalists who had already known our work from Syria trust us already. So we already were getting attention early on by people who had that trust in us, in the journalism world.
But who was we in that period? How many people were there.
On July 14th? It was me.
Yeah, it was just you.
And then a few days later. I mean, these communities online formed instantly around what happened around MH17. And I could see right away that there were some people there who were good investigators. And so we kind of were direct messaging each other. One of them, Aric Toler, who now works for The New York Times and was working for Bellingcat for several years. He was just someone who saw one of the images and said, oh, I think I figured out where this is. I asked him if he would like to write something for Bellingcat. He said no, but he was happy for me to write it up myself and that's what I did. But eventually he became part of our kind of little volunteer community, and we're talking like 5 or 6 people who are like really close knit. And when I had that interview with the police, I suggested we all join the same kind of Slack server so we could coordinate a bit better. And that then became in October, the kind of, I guess you could call official Bellingcat volunteer investigation team.
And how did you get funded in that period?
It was crowdfunding on Kickstarter at the time. Yeah. So I raised, I think about it was about €50,000. That was unusual because I launched it on this Kickstarter. And it wasn't going very well at all. And then, Hugh Grant, the actor who had been, I had done a lot of work on the phone hacking scandal. He gave a big donation to it, which kind of pushed it over the top at the end. So weirdly, Bellingcat probably wouldn't exist if Hugh Grant hadn't put that money in.
That's a good story.
Yeah.
That's a very good story.
And nowadays you are based in The Hague.
In Amsterdam.
In Amsterdam? Yeah. And is that a how should we visualize that? Is there a is there a newsroom in Amsterdam?
Oh, no, there's not like a newsroom with. I think people picture us out with kind of six monitors in front of us and.
Yeah.
Lots of server lights blinking in the background. But, that's really just a kind of administrative hub because we now have about, I think it's about 37, 38 staff members. So there's a lot of administration involved, especially when you have people working across the globe. I mean, one of the most complex things we do is in our investigations. I would say it's actually our administrative work because with so many tax authorities to deal with and all that boring stuff. So, but yeah, we're all over the world. I mean, I'm based in the UK myself. There's a few Bellingcat people there, but it's rare. We kind of hang out with each other.
So it's an online newsroom in a way.
Yeah. Very much. And but it's always nice when we get together at staff retreats and other events because we a couple of years ago, it was ten years of Bellingcat. So we had a two day festival in Amsterdam in a theater where, it was lovely to see people from across the whole community come together in the shared space. There was a real unique energy in there, lots of people sharing ideas excitedly. It was fun.
And what's the link to the Netherlands? Why?
Well, I wanted to. So the first few years of Bellingcat, I wasn't really sure what Bellingcat would be like. How do you do this as a business that can actually, you know, make money and, you know, pay for my mortgage basically is my concern. So, but it came clear and clear that a commercial model wouldn't work because of the kind of process we use for investigation. Bellingcat with loads of advertising on the website wouldn't work. We didn't get that kind of audience. We had a big volunteer community, so kind of turning their work into a commercial project. Project to me seems like wrong. So I very much wanted to go down to the kind of nonprofit route. And in the UK at that time, you couldn't have a like be a journalism nonprofit effectively. But because of our reputation in the Netherlands, we already had a lot of support there. In 2018, I registered Bellingcat as an ANBI. So I ran it as a non-profit and, yeah, I mean, that's been good in one way because we've been quite good with funding in terms of being independent of what we can do. We don't have to chase commercial pressures.
But can you tell us how you operate? So, so where does the money come from? You have 37 staff members. You pay them probably from that is there enough funding? How does it work? The concept of Bellingcat.
Yeah. So, generally our funding has been split in a few ways over the years. About 30% of our income we can generate ourselves. So we usually do that through workshops. So we've probably trained in workshops longer than two hours, about 6,000 to 7,000 people over the years as journalists, civil society actors, and we also run private workshops and stuff like that. And then we have, various family foundations that donate to us. We have, postcode lottery. So the Dutch postcode lottery is a huge supporter of us. And yeah, they've really meant a lot to us in terms of their support. We've been sponsored by the Swedish Postcode Lottery, for example, and then the rest comes from various small donations. But we are planning soon to launch a, supporter programme where we'll ask people for a set amount of money, kind of following the model that's been successful for people like follow the money. Yeah. So because you do have to really diversify your streams of income now, because you just can't rely on one stream of income. And I think having, kind of supporter communities now is something that's really important for any independent NGO and the US funding cuts. You know, Musk's and Trump's cuts have been devastating to the sector globally, it's been estimated because of those cuts, there's been a drop of 25 to 40% of funding for independent journalism. And the really, the journalists in the Global South have borne the brunt of that. But it has an impact on the whole sector. So it's a very, very competitive fundraising environment at the moment. So that diversity of funding revenue, is something that's essential, I think, for the survival of any NGO now.
But how do you pay your staff members? Is it like a full time job salary? Because yesterday I watched the documentary on Navalny and your lead investigator on, I think Russia, Christo Grozev.
Yeah, he was, basically it was his hobby that he was paying for.
But there was this, there was this one scene where they asked him, how much money did you already invest in your, in your investigations? And, they asked him, is it a couple of thousand? And then he said, no, I think I put already in like 150, zero €0 of my own. But don't tell my wife. My wife doesn't know this. And then the interviewer said, but she knows now. No, she's not going to watch it, so I'm not worried.
The funny thing is, I was sat next to his wife when we were watching that at the premiere, and Chris, Chris, I could feel Chris kind of getting really worried about what was going to happen. Fortunately, he got a standing ovation, which I think kind of helped a lot. But yeah, it was probably there was an energy I felt from that direction.
There was a funny scene.
Yeah. And Christo was also targeted by a Bulgarian group of spies that were, arrested and, condemned in, in the UK, I think.
Yeah. I mean, even the background of that is weird because the Bulgarian spies themselves were being run by a guy called Jan Marsalek who was responsible for the Wirecard fraud in Germany. He basically disappeared with, €1 billion from Wirecard, a payment provider.
€1 billion.
Yeah.
And it's a lot of money.
And disappeared. And Christo actually tracked him and showed that he had flown to Belarus when he was trying to fake his kind of grew out of the country. And it turned out and we discovered this during the investigation that the police were doing into this spying ring, that he was running this ring and he was running it as a proxy to the FSB, the Russian intelligence services. The thing that was ridiculous about this is the ringleader of the Bulgarians. And Marsalek had a signal group and they didn't have disappearing messages on. So over this, you know, two or so year period of spying on Christo and other members of Bellingcat, he they just had like, I think it was like 60 or 70,000 messages just going back and forth, giving a blow by blow account of everything they were doing. Like there's one conversation where they're talking about breaking into one of Christo's flats in Austria, and they're like messaging, oh, we've listened to the door and there's no one in sight. And when the police dusted the door for fingerprints or prints all over the door from them listening, they went in and stole a laptop from it. And then there was like two weeks of them going back and forth about smuggling the laptop out of the country. All these convoluted plans and routes they were using like this was like the most precious thing ever. And they got it out of the country and they got it to an FSB lab. And Marsalek was talking about that. And then he said, oh, they've come back to me and they've basically called us idiots because there's no hard drive in it, because it was all Christo's old laptop he'd just abandoned, and they'd spent two weeks smuggling out of a car.
No way.
Yeah.
So but the whole thing was documented in the conversations.
But were they sent by this billionaire, you think, or were they, instructed by the Russian intelligence services?
I think Marsalek is very, very close to Russian intelligence services. I don't think he is an intelligence officer, but he's kind of like a useful proxy for them. And, you know, that gives them a degree of deniability. Assuming you don't have a 9000 message signal conversation that you've left all the evidence of your activities in.
So you are you are actually doing dangerous work in a way. Do you have precaution measures and.
Yeah, I mean, it comes in, you know, various levels. I mean, there's constant kind of cyber threats against the organization. You know, there's disinformation campaigns that have been run about me and Bellingcat state media, for example, in Russia, we know obviously we've been spied on as well. So we have a very kind of security conscious culture in our organization. But on my personal level, you know, the local counter-terrorism police in the UK have people who come and visit me every couple of months to check in with me, see if we've published anything controversial recently. Let me know if there's anything that's kind of bubbled up. I think they must have a connection to Gchq in the UK, because occasionally I'll get a letter about a threat from a state actor that usually it's around going after dissidents, so it doesn't really matter to me, but they let me know if something bad's going to happen. So that's nice to have.
Yeah. So, I actually heard you saying, that, that you don't, you don't order room service, food in the hotel.
Yeah. Because it's even with the best intention that's always in your head that there might be something strange going on. It's like we had a staff retreat, and, on the last day when I turned the air conditioning on, there was an awful eggy smell, like a really unnaturally strong eggy smell that came from the air conditioning. And it gave me a little bit of a panic attack because it was something unexpected and unusual. But then, you know, I didn't die. So I thought, okay, maybe it's just a dirty air conditioning. But then we found out later the Bulgarians were in that hotel spying on us that week. And there's even a photo of one of them took of me and Christo having breakfast one morning. So now I wonder, did they actually put something in that air conditioner, even if it was a stink bomb just to scare me? Maybe that was it. Or maybe sometimes my just my own kind of ignorance towards those risks is something, you know, keeps me calm and, you know, safe.
What I thought was interesting in the documentary Navalny was that, Christo did his research. He tracked the people who were working in the, the factory or the spot where they made the nerve toxin Novichok. And I think if I remember it correctly, they, got their phone numbers and based on their phone numbers, they got their names and then they checked if they were coincidentally flying to Tomsk in Siberia, and it checked out like 3 or 4 people were in there. And then he sent like a message to Navalny through Twitter, probably a DM like, we, I think I discovered who tried to kill you. And Navalny said after that he didn't really trust it because there was this one guy from Bulgaria who just, explained to him the whole plot. So it was not necessarily like the concept of Bellingcat behind it. I, I got the feeling that Navalny didn't really know what Bellingcat was at that moment.
Yeah. And I think if you encounter us in the wild and we're talking about all this kind of stuff in detail, it's weird enough. But when someone direct messages you and tells you, I know all the details of your poisoning. Yeah. Would you like a chat? I mean, we get a lot of emails like that into our inbox, and we normally file them in the kind of crazy folder. So, you know, building that kind of relationship with people is important when you're explaining this stuff. And this is partly why I do a lot of, you know, talking about this work, because we also want to explain that a lot of this work is something that anyone can really do. Yeah. Now Christo, I will say, was using kind of this black market of data in Russia, which now we wouldn't do because. Yeah, because.
That would be my other question. How do you look at that? Because Christo, he bought on the black market like data from probably Russians who didn't, who didn't, who do not earn like a lot of money and want to have a sidekick. Right. And he kind of, uses this vulnerability, I guess, in the system of Russia to get to the data. That normally is, is obsolete to him. So now he pays people from the black market and gets his data. How do you look at that?
Yeah. So, at the time, I mean, we prefer to use purely open source stuff. And it was almost a kind of something we got drug into a bit in terms of doing more and more of this because when the, so this really started, I would say with the Skripal poisoning. So, there, when the poisoning happening that happened, there wasn't much information about what had happened a few months later. The identities of the two suspects were released and a Russian news website, purchased the flight manifest of that flight and found their names on it. And what was unusual is the two passport numbers were only a few digits apart, which is very unusual for people who don't appear to be related to each other. So, we were kind of discussing this on our Slack and Christo said, well, I know a way to get their kind of passport registration documents. So this was for what's known as a domestic passport in Russia. So you can fly between the parts of Russia that are kind of separate from each other. And, so Christo basically went to one of these kind of black market people. You can find them in the internet. There's just, there's so many of them because Russia's corrupt basically, and got this form and we didn't think it was going to have something like, you know, Russian Spy stamped on it. But it's pretty much what we had when we got it because there's loads of information missing about their past identity.
There was also a stamp on it that had the phone number of the Russian Ministry of Defense that basically said phone this number if you see this document. There's another stamp on it. Which basically says Secret Service on it. And the thing is this was just as they were on like literally as they were on Russia today saying we're tourists going to see Salisbury Cathedral's 123 metre spire. And we were looking at these documents that obviously show they were fake identities and Russian spies. So then Christo was like, well, I'll buy some more stuff. And at every point we expect it to fail because we thought there could be no way they would just leave this ridiculous amount of evidence about the identity of their spies online to be pieced together. But that's what there was. There were people selling, you know, registration databases to like that, where they were living in Saint Petersburg and how they and Moscow, where they just weren't there on the previous year's database. And then they appeared on the new database the year later showing that these identities had been created. So that kind of became the rabbit hole. He went down and there was so much there to find. And then Navalny was poisoned. And that same process could be used to identify the poisonous. But as time went on, we were really concerned this was less about open source investigation. And we really wanted to double down on open source investigation as an organization.
Because in the beginning, Christoph didn't believe that he could do some kind of investigation into Navalny and his poisoning because it was all in Russia. And he thought, okay, because he didn't even try or Bellingcat didn't even try. And then afterwards he came, well, maybe I can do something with the information or the data on the black market. And also what was interesting is what he said is, in a time of fake news, basically he said, I don't trust real people anymore, but you can't trust data. What is your, what is your view on that?
Well, I think you need to be careful about the data. I mean, one thing we did find when we were doing that investigation is we published that story about those passport forms. And then when we identified a possible third suspect, we tried the same trick with him. And when we got that form back, the photograph had been removed from the form. But then that told us that there must be something interesting about this form. So we published about that said the photo is removed. That seems a bit odd considering they should be there. And then we did another person like that and this time the photo had been changed. Obviously to try and cover the fact that, you know, removed the original photo, but for some reason they changed it from a photo of a man to a photo of a woman. So I don't know why the woman actually funnily looked very similar to the man. So we wonder if they were using some like weird facial recognition system or god knows what. But it was just we kept on coming back to kind of finding like little clues that connected. I think in the Navalny case, what it was is the guy Christo identified as being the kind of chemical weapon scientist who was involved with the Skripal poisoning.
He got his phone records, and then he had all these phone calls around the time of the poisoning that Christo looked into the numbers of. And they all turned out to be working for the FSB. Yeah. And keep in mind, this could be as simple as using a. So. So you have these apps, in Russia that are like phonebook apps. So you meet someone, you put their details into this app, but then it goes to a centralized server. So when someone else gets the same number, it gives them that, that information already. But people would make, you know me, Igor from the FSB, and put it in their phone as Igor FSB. So there's bots that allow you to basically query these databases and you get all the names that are saved under. So Christo would put the name in and it would come back with like ten names of like, you know, Igor spy, FSB, Eagle, you know, it would be like four different or five variations of that same thing. So that helped us identify who the spies were. Quite helpful.
Eliot, this all sounds as a kind of naivete of the beginning years of open source investigation. Isn't it more difficult nowadays because FSB is not that stupid anymore. And people and sources are open. Sources are being suppressed or closed or even or not used. And also soldiers, for instance, they are, in on the Ukrainian Russian front. They are said, don't use your social media, don't use your smart smartphones are taken from them, to prevent this kind of information, being open source.
Yeah. I think to one extent, when they, I mean, we've taught Russia a lot of lessons about open source investigation, I think it's fair to say. So they have done things to mitigate against that. So the Duma passed a law to actually make it illegal for soldiers to post information about their military service directly because of the work we were doing. But that's just Russia and Ukraine, another place in the world. They don't have a clue. There's an investigation we did into, a really brutal murders in Cameroon by soldiers who were operating there. And when we initially published that investigation with the BBC, the Cameroonian communications minister gave a press conference with a still of the video behind him with the words fake news written underneath it. But I think within a year, all those soldiers had been arrested, put on trial and convicted for those killings. Because I think I also heard anecdotally when, those Ukrainian plane shot down over Iran a few years ago, flight PS752, shot down by an Iranian air defense system. And we quickly found the video that showed, at night showing the kind of missile streaking through the air and hitting the aircraft. We geolocated it showed it was the flight path. And it was clear evidence that this was a ground launch missile that had done it. Iran, of course, was denying it. But when the Ukrainian investigators went there, the air crash investigators, they said Bellingcat has already said what's happened. So don't even bother lying about it. And like within days they'd admitted, yes, it was us and that we did this.
So I mean, we've had that another level as well. We did this investigation one into once into kidnapped dogs. Yeah. Because during the Covid lockdown, dogs were being stolen from parks when they were being walked. And we, this UK organization called Doglost contacted us and they sent us this video of this car speeding away from a CCTV camera and asked us if we could decipher the number plate. And it was blurry. It was hard to see, but one of my colleagues had developed a system to, using machine learning algorithms to decipher blurry number plates and use that in an investigation into the Russian journalist in Ukraine, Pavel Sherman, because there was a CCTV footage there. So we applied that same technique within 15 minutes. We had the number plate. We matched it to the make and model of car to make sure it was correct, pass that information on. And the police went to the house and they found the dog in the garden, and they picked it up and took it back to the owner. We did another one like that, which was very successful. But the third time they came in touch with me and said, hey, we just want you to know we've had another stolen dog. But this time we just sent them the results of your last two investigations, and they just gave the dog up straight away. So it has this kind of preventative effect of kind of using open source. It scares people into behavior sometimes.
So Eliot, what does your methods mean for journalism? Should we change? What, what, what are we fighting nowadays? And how can, how can we use open source investigation?
Well, I mean, a big part of our work over the years has been training other people to do this. This has never been about what can Bellingcat do? We don't see ourselves as the center of this universe. So a lot of the work we've done has trained journalists who have worked on things like, you know, the conflict in Ukraine, for example. And that's why you've seen such an increase in the use of open source investigation. I think particularly since 2022. But I think there's something kind of broader here. I think there's a more deeper structural issue that we have really moved into a different information environment as a society compared to where we were in the 20th century. If you think about media models in the 20th century as very much top down, elite controlled and gatekeepers, it was very procedural and structural when it was working at its best. And the public in that model were a recipient of information. And they were also in an environment where there was a limited sources of information. There were only so many newspapers, news channels and so on, and there was a kind of excess of attention for that information. You had a public. So this is a very stabilizing time because, you know, the people who fought the earth were flat, were kept off the television and out of the newspapers for the most part. But now we live in a completely different environment where we have this many to many peer to peer environment, which is algorithmically mediated. So what works really well, there is stuff that's reinforces people or entities that has an emotional element to it, stuff that's just engaging. Sure. So that incentivizes people to share and create that kind of information.
And the public as well are not just recipients of information. They are participants in its spread and creation as well. So this is a very different role for the public in this current information environment. And it also is a different role, I think, for institutional media and other institutions. Because the question is, you're no longer at the top of the pile kind of putting that information out there. You are part of a kind of network now, and I think the media has to examine its role in that network. And, the work Bellingcat does, a lot of it is not just about doing the investigations, not just about the education side of it, but building communities and spaces where people can come together. We have a discord server with 42,000 people on it who get involved with investigations and come up with their own investigations. We've built university hubs with places like Utrecht University, where students learn how to do open source investigation in their courses and then apply those skills in an investigative hub. In the long term, we want to build more of those and connect those to local media, local community, civil society actors to collaborate with them, using those open source skills to investigate things that matter to local communities. And I would say I'm a very, very strong believer in local media and their connection to local communities being part of the solution to this problem we're facing, where everything is so networked, so peer to peer that allows us to kind of work around the algorithm. If we have these personal relationships with people and we're doing more of a kind of public interest news with local populations.
But it's also it is also your view that, people should harness themselves, against this new information area. Is that a correct way of seeing it, that people should learn how to investigate themselves or how to value the information that gets to them?
I would say yes, but it's kind of how do we approach this? This isn't trying to make everyone into a kind of world of kind of open source investigators. You know, people don't have the time to do that. I think one challenge we have is that because the public now has this role that, you know, the institutions had in terms of verifying information, they acted as this procedural filter, which is now in the hands of the public. We don't teach them and educate them around that. We don't really even explain to them why they have this role. Now, the importance of that role in a networked society that you liking a social media post might seem like a small act, but when it's millions of people doing that, it informs an algorithm that brings it to even more people. It creates this kind of cycle that people can kind of be drawn into to an extent. So we have to educate people around that. And that's not to say you've got to teach kids how to fact check headlines in school, partly because they never see headlines anymore because they're on TikTok all the time. But, build those critical thinking skills they need because, you know, a young person nowadays, they might get their first smartphone at 11, 12, 13 Be on social media. And we kind of assumed they just have the skills to, you know, they're all like in, you know, inherently media studies students, like they know how to translate this media, they know what's going on. But I think we need to build education around that. And that's also part of what we're doing at Bellingcat, working with primary, secondary educators to build curriculums and teaching material around this, especially when teachers are increasingly becoming almost frontline social workers for a lot of young people as well, and often encounter these conspiracy theories or ideas they're seeing in these spaces. You know, they're the first adults who kind of encounter that in interactions with children. So they almost have to learn how to understand this kind of thing themselves.
Yeah.
You said that you would not define yourself as being a journalist, not ten years ago, and even not now, how would you define yourself? Is there is there like a definition to give from Eliot Higgins?
I, don't know. I don't want to give myself any grand titles, that's for sure. Yeah, I mean, I do. I guess you could say acts of journalism, but a lot of the work I'm doing at the moment, I mean, there's the administrative side of Bellingcat, which is one thing being a, you know, a director has its own responsibilities, not all of which are terribly fun, but, you know, a lot of the work I'm doing is getting out there talking about our work. You know, thinking about these ideas, building these communities and these coalitions and finding different types of partnerships. So whatever title you want to give, maybe there's a Dutch word for that, I don't know, but it's kind of I'm just out there doing my thing really.
But is there like a centralized way of working for Bellingcat right now? Because how do you how do you choose your topics?
So, we have different investigative groups within the organization. So like a financial investigation online or ideologies, environmental investigations and conflict. And each of those teams has a team leader and a lead editor for those teams.
And those teams are not situated somewhere in the country. They're also situated all around the world.
Yeah. So they're spread all over the place. And then they're looking for stories that are, you know, leads that are potentially interesting and they'll start digging into them. And then they start working with the editor to think about, could this be an investigation or is it not going to go in that direction? Yeah. But also, I.
Mean, like, like some topics are so easy, like, like MH17, obviously that's a topic, but like you mentioned Cameroon, for example, that's, that's a, like a small country, somewhere in Africa. It's not, it's not completely, the first thing that comes to mind.
Yeah. I mean, I would say there's certain things that are more opportunistic, I would say, in terms of seeing something that is kind of going viral on social media or there's a lot of discussion about, it's like recently with the Alex Pretti and Renée Good shooting in Minnesota. We did a lot of work there where, we understood how the discourse was going there, that it was a lot of, you know, one side versus another side. It wasn't really about the evidence, it was about opinion. We knew the government was going to lie about it as well, because they've got a good track record for that now. So, we wanted to get information out quickly. So rather than, you know, slowly investigating it and get something out in 24, 48 hours, as we verified kind of individual bits of video, for example, we would produce a bit of content to show on social media within 2 or 3 hours of the shooting happening. So there's a debate about whether Alex Pretti was disarmed or not, if he was carrying a gun. Yeah. So we had, 2 or 3 videos. We synced to show multiple angles of the gun being removed so you could see very clearly the gun being removed and then him being shot, and that he was never holding the gun in his hand. He wasn't reaching for it, that all these kind of little excuses people were coming up for, the behaviour of the officers could be eliminated as well. But because the discourse was ongoing and, you know, it's not just ordinary members of the public involved of that. It's the journalists who are going to speak to the white House about this as well. They could see our work and then be confident that it wasn't like an open question then.
So is that a way that Bellingcat, influences media? Because you obviously don't have the mass visits to your own website that, that other big media have. But you also work together with newsrooms around the world.
Yeah. So that collaboration is very important for us. I mean, it's always been part of our work, going back from when I first tried to connect people together when I was blogging. But, I think one thing that's important now is because we've had this fracturing of audiences, you often have to go where the audiences are. So, we did an investigation recently into the Kinahan drug cartel showing that their leadership was hiding out in Dubai. But when we published that, we partnered with a YouTube channel called Search Party. Who does kind of form, you know, ten, 15 minute kind of short documentaries, really, really good graphics for an audience who is interested in drugs, sports and crime. So, you know, that's a tick, tick, tick for what we were talking about in this investigation. And you know, their video I think is now like 2.1 million views, which is way, way more than we could expect to get on the Bellingcat website. And I have people now who talk about that investigation to me that they saw on that YouTube channel, not on our website. And we don't need people coming to our website. We don't look at our numbers and pat ourselves on the back. It's more about how do we reach those audiences in the spaces they already are.
So that's part of our thinking when we do these collaborations. But also we're working, we did something, for example, in, immigration raids in California. And there we worked with CalMatters, a local news organization. We worked with Evident Media who produced this high quality YouTube videos and other content, and we brought our open source investigation skills to that. And then that little coalition was able to have far more impact than just CalMatters working by itself because we had this interesting visual content, we had the open source content there on the ground investigation, and we've done other investigations like that where we work with a national news partner, like we did one with a PBS frontline, for example. And you know, this like it has it's recognized, it has a huge impact. It's like we're currently nominated for that work for, I think it's six news Emmys, which is all collaborative work. But I think it's, I think it's I mean, in one category, we've got two different, nominees in it. So I think it shows how powerful this work can be. When you work in collaboration, you think about new ways of doing investigation, but also drawing on those kind of more traditional and new media.
You also influenced by teaching us that transparency leads to trust because you always, from the beginning, you say how you could access the sources yourself and actually repeat the open source investigation. And this is not a thing that journalists used to do, and I would just not show how the sausage is made. But now nowadays we try to do it more in order to gain trust.
But it feels like it's like we still don't do it. Like not the way Bellingcat does it. Because you mentioned the case of Alex Pretti in America and you would post, like a new founding immediately, right? And step by step and show the process. But if, if like a journalist from the New York Times would investigate this, he probably would finish or wait until the whole project is finished and they have a decisive, answer to tell. Right. So, I figured that this is like a way between how the media works or the news media and the other way how Bellingcat works. And in your opinion, should we be working more like the Bellingcat method? Should we adopt the publish immediately version?
It's I think in a way, it's you can look at it in terms of discourse and these modern information environments, because opinions are made and shared very quickly. And there's a demand side to this when there is one of these kind of high salience events on social media, you have lots of people looking for information about it, and you have plenty of people who are willing to provide information, but it's usually opinion or spin or something that supports their own position or is appealing to a certain audience. So in that moment, you kind of have to question what's the role of open source investigators? Because what they are finding and discovering can contribute to that discourse. But they also need to be confident they're sharing accurate information. And this is why the Alex Pretti shooting, we weren't saying this is the whole story of the event. It was actually focusing on one element that was being discussed and saying, this is the evidence that's available so far. Here's a presentation of it in a way that makes it easy for you to see, but that also becomes a kind of transferable object. People were downloading that video, sharing it on other platforms, using it for their own online arguments and fights. But it was adding something that was, you know, verified information to an environment where people were seeking, you know, any information.
But I can I can see this in a, in a traditional media outlet as well. Like, for example, a live blog. Yeah. And then, and then post your next finding.
The thing with a live blog is often that's kind of, you know, on a set that's outside of this discourse. It's something people have to find and discover when we're kind of in that discourse. It's not just about us being kind of one point information radiates out from it's people kind of taking those objects we're adding, adding to it and distributing it themselves into different platforms. Like one thing that really surprised me is within a few hours of the shooting, we had quit X because of the non-consensual nude stuff it was doing and other many other issues with X.
Because as fascist probably.
Yeah, that is a small problem as well. But we also, what I found surprising is within a couple of hours, people were asking Grok already who's responsible and stuff like that. And it was already citing our work because it had kind of scanned it within the kind of people discussing it on its platform. And Grok really trusted us. For some reason. Elon Musk has called us a psyop, but Grok trusts us. So, so immediately it starts propagating into these AI systems that are kind of reading this data through kind of, you know, the public. And you also have to think in terms of technology. Now, the public and the kind of technological platforms are kind of so intertwined now. They kind of almost interaction that kind of how we've been consumed by the technology is the kind of substrate, the kind of medium for the discourse around any events that happen like that.
And speaking about technology, does the rise of AI, help your work? It also makes the environment more complicated and more fake images and so on.
Yeah, I definitely have a love hate relationship with AI.
I can imagine.
Yeah. I mean, it's, I mean, in terms of how if it makes our job easier, we've tested them on things like geolocation and they give a very wide quality of results. I would say sometimes they're very good, sometimes they're very bad, but it can kind of produce little bits of information that you may not have noticed. Like you stick a picture in and it says, oh, that tree is only in this region of the world. And you're like, okay, I'll double check that. But I hope so because that helps my job. Sometimes it can say, this is definitely this location and it's completely wrong, or it can be completely right. So, you know, it's a mixed bag.
So it speeds up your work. It can do investigation.
But it gives us an option. It's a tool. It gives us an option. It's always good to have lots of.
Which you always have to check afterwards.
Yeah. Then I think there's a lot of concerns about, well, you know, will these images affect your investigations? There's always been misattributed and faked images. So we have ways of mitigating that of using things like geolocation, multiple sources of data. We don't just rely on one photograph. And this isn't the time really to do that. I think though, that one of the biggest risks is it gives people a kind of permission structure to disbelieve true information, because now they can say, well, this is fake. This is AI generated.
I don't really see this happening, you know?
Yeah. You see that a lot.
But they could already say it. They could already say.
Yeah, they could say that. But now they've got even more of an excuse to say that. And they don't even have to say, oh, well, you can't trust this group or you can't trust this guy. That's just AI. Yeah.
And but we also see the possibilities, for instance, in our investigative journalistic work of vibe coding. So accessing data with a, with a good prompt, that, that, that, that were not so available before because you didn't know they exist and AI finds it for you.
Well, yeah, I've used some vibe coding myself for a couple of like small things I've been trying to do. Like when I've got an annoying file format, I want to convert automatically rather than paying for some, you know, ridiculously expensive piece of software. I can just find something that does it for me in five minutes. So I think there's value there. I would actually say if you're someone who's a coder, I feel quite sorry for people who've spent all that year training because the power I use Claude, for example, and I'm amazed at how well it can code even fairly complex processes. And they actually work reliably.
Yes, but it still.
Helps you if you are like a qualified coder.
Oh yeah. Absolutely.
Because you have to check it afterwards. If it's really doing the thing that it says it does.
Yeah. I think when it gets increasingly complex, there is that those skills that are needed. But the advances we've seen in terms of coding, I think from like 2 or 3 years.
That's crazy.
Is I mean, obviously there's been lots of advancements, but that's one area I think, it really does have an advantage. And of course, now you've got these concerns about Anthropic's Mythos system that can hack computers. Yeah. Mythos that can apparently hack computer systems. And God help us. In fact.
It could also be a marketing story.
It could be. Let's hope it is.
Because maybe you can find out.
It's like saying, we've invited a brilliant new robot that can travel through time and terminate people. And that's our advertising campaign. Maybe that's something Elon Musk is going to advertise at some point, but I.
Are you training your staff for how to use AI? Are you motivating them to use AI? Are you giving enterprise, subscriptions? How, how do you approach this?
So yeah, we do have an enterprise subscription. I think it's for Google's AI system. That's partly to experiment with it to see what works and what doesn't because a new version comes out and then you start mucking about with it. Yeah. but we have a lot of rules around in terms of using it as a kind of writing tool that's not allowed. If it's used for any part of research, it needs to be explained exactly how that brought us to it. If it creates a conclusion. That's not what we're interested in, we're in. If it discovers evidence and then we can say, here's the evidence online, then that's something we can link to as an open source. But if it comes to a conclusion, then that's not what we're looking for.
Can I ask you maybe one last question? If you look at the media field, if you look at how editorial, departments are working, how they are investing in investigative journalism and maybe an Osint techniques, would you say they, should be doubling down on it? Do we do we investigate enough? What is your view on how, journalism is incorporating Osint in their way of working?
Well, I think one thing they need to think about is it's not just about your staff and what your staff can do for you. I mean, it's great if you have lots of open source investigators, you know, if you're like the New York Times and visual investigation team, that's fantastic. You'll do great work. But also it's about community and, you know, what are you able to build in your organization that connects you to a community? Because once people have that sense of belonging, they, you know, they contribute to it. We find this all the time with our discord server on Bellingcat. It's not just a nice place for people to come together, but they do investigations that we're able to pick up, develop with them, and then publish with them on the byline. There's, you know, I think a lot of people who are out there looking for serious answers to these problems, these issues, who aren't interested in just having another person give their opinion from their kind of news influencer YouTube channel, whatever it may be. So I think, I think traditional media really should look at a way not just to engage in an audience in terms of, let's have a comment section, but actually see if there's a way they can build public participation in what they do. And I think again, at a local media level, that's something that's really, really important. The idea of public interest media collaborating with the public is something that I think we need to really think about.
What kind of stuff could I do then in the local journalism?
Well, I mean, this is why we're working with these university hub model that in the UK, for example, we have the Public Interest News Foundation. There's been a real loss of local media in the UK, but a kind of growth of independent media organisations in cities and regions in the UK. Those are exactly the kind of organisations you want to work with from a kind of open source perspective, because one tool we built recently was a tool that allows people to look up transcripts of council meetings from across the UK, which didn't exist in one place before. So those tools are now useful at that local level, especially now we are having more reform. Councillors turn up who clearly don't have an idea of actually how to do politics, because they've spent all their time on X being racist. So that allows people to actually track what those people are doing, rather than kind of and kind of being been outside of the conversation, which is often the case. And by making those tools available to both, you know, the public and those more institutional actors, there's an opportunity for the public to use those tools to find interesting things. So then the question is, where do you have that interface between the public and those other actors? And for us, it's, you know, having discord servers, for example.
Quite interesting.
All right. Thank you so much. Thank you very much.
For coming here, Eliot.
This was the Media Innovation podcast with our guest, Eliot Higgins from Bellingcat. And obviously my co-host Philippe Remarque and myself, Lars Anderson. Blijf luisteren en kijken.