The power of digital crowd

KATARYNA PRYSHCHEPA: How did Bellingcat come about?

CHRISTIAAN TRIEBERT: Bellingcat was started by a British man called Eliot Higgins in 2014. It evolved from his blog which he had under the name of Brown Moses. In his blog he started with analysis of cases of weapon use in Syrian war and contributed to the comments section in the Guardian under the same name. More and more people started to respect his work, including human rights organisations and respected journalists. In 2014 he decided to reveal his identity and start a collective – an open source platform called Bellingcat, which would coordinate digital open source investigations. The aim was to investigate basically everything, be it the downing of MH17, illegal tiger hunting or chemical attacks in Syria. Bellingcat was founded with a crowdfunding campaign. Several days after it was created, the MH17 flight was shot down.

So at the beginning Eliot Higgins and Bellingcat were mostly concentrating on Syria?

Yes. It was mostly Syria. But when the MH17 crash happened people started sending us different stuff to look into. Mobile phones usage in Eastern Ukraine is high, so there was a lot of material. And until now the MH17 is still one of the biggest among our investigations, but not the only one. At present we have investigations about Georgia, Montenegro, weapons from the Balkans – this is the case which I am personally investigating right now – Iraq, Yemen, Central African Republic, Colombia and so on. Thus we cover a lot of different locations. Sometimes journalists do not have a possibility to travel to an area where presence on the ground might be needed. Therefore, we are trying to encourage people to do the investigations – to the extent it is possible – digitally themselves and teach them how to do this.

And how do you teach investigating techniques?

We have a very small team and not a massive amount of funding. Bellingcat is not a traditional organisation. We all started as volunteers, as people who were just very interested in this kind of work. Right now we only have six paid colleagues – including me – and all of us are freelancers. In order to make a living I also do other work – writing for magazines, making presentations and teaching at different workshops. In addition to this group of six paid contributors, we have a core group of 15 to 20 volunteers and all of us are in a daily contact with each other. So in total, Bellingcat investigative team is about 25 people. Volunteers cannot work full time. For instance Timmi Allen, who identified the car plate in the Pavel Sheremet investigation, is a volunteer. Because we are such a small group, we cannot afford extensive in-person teaching. But we did workshops in a number of countries, including at the Kyiv-Mohyla Academy for their journalism students and Stop Fake invited us to do workshops for them. I also was invited to teach at media institutions in Iraq and Colombia. We mostly teach journalists.

Do you also teach policemen or the military? They might be interested in this kind of techniques.

I can understand these techniques are interesting for the police or military, but I think it is best to avoid such contacts because Bellingcat is investigating precisely the activity of the military.

If you were approached by the military or police to conduct a workshop for their personnel would you accept it?

I strongly feel I would reject the military. And I have done so in the past already. Teaching the military may not be a right choice, because if I have to investigate them in the future, I do not want to have a conflict of interest.

The police are a bit different. In MH17 investigation we provided the Joint Investigative Team, which consists of representatives of the Dutch, Ukrainian, Malaysian and Australian police, with materials from our investigation which could lead to potential suspects. And for that reason we did not publish all materials on that investigation.

So you did provide the results of your work to the Joint Investigation Team (JIT) but you did not conduct any technical workshops for them to explain your techniques?

Bellingcat provided the results of our investigation to the JIT and Eliot Higgins has been interviewed as a witness, mainly going through Bellingcat MH17 evidence step-by-step. Bellingcat regularly sends the JIT information and unedited versions of our reports, we try to refer witnesses that come to us to the JIT directly and ask them to give the JIT all the information. Bellingcat thus provides the JIT with information, but this is a one-way stream of communication.

What is the background of the Bellingcat team? Do you need to be a computer geek to do this kind of work?

I personally do not have any special computer skills. I do not know anything about coding or programming, for example. I think very few people contributing to our investigations have such skills. I can name one person contributing to Bellingcat, Justin Seitz, who has very high computer skills and develops his own tools. He is a part of Bellingcat group and but he also does his own work. 

It is very important to realise that Bellingcat operates in an untraditional manner and we really work collectively. Only six of us are getting paid and of course we are trying to coordinate our activities and discuss the priorities. But in the volunteer group you cannot really tell people to do things. They choose themselves. Some of them have good computer skills and others do presentations and geolocation, for example, and this is really Bellingcat’s strength. 

And what about the language skills required?

This is a very important part of our work. We have, for example, people on board who speak English, German, Dutch, Spanish, Russian, Ukrainian. And we have a network of people, who can help with translations from Turkish, Urdu, Farsi. I myself also studied Arabic, though it is currently not sufficient to properly speak it.

Do you rely on computer translation?

It is better to have a person who can translate things, but computer translation is sometimes very useful.

Do you think about the fact that IT companies can use your experience to develop new tools to be used by state actors, including intelligence services?

The very name Bellingcat comes from that legend about the mice which put a bell on a cat, so it could hear it. We always try to be open in explaining how we are doing it. But on the other hand because we are open about the tools we are using, the “cats” which we are investigating can try to use similar tools. We also we have to be careful about our funding. For instance, my position at Bellingcat at present is funded by Google’s Digital News Initiatives (DNI). So we have to make sure that we can have independence even if we are receiving funding from big institutions. Fortunately, that is the case with the DNI funding.

Since you mentioned your funding, can you tell us how Bellingcat is funded now? Do you continue with crowdfunding?

Yes we continue our crowdfunding campaign and we currently have the funding from DNI, which I mentioned. Previously Bellingcat also secured for funding from OSF (Open Society Foundations) and Meedan, for example.

So far the MH17 is the most famous of Bellingcat investigations. Can you tell us a bit more about it? How would you say the work was divided between the Bellingcat team and outside volunteers?

It is hard to say really. At present there is a real big community, especially on Twitter, of people who discuss the issues of this sort. And sometime people would just send us material. So we really relied on the power on the crowd in this case. We invited people to check the authenticity of some evidence or its geolocation. We used this technique in case of Syria and Iraq investigations too. It is hard to say whether it was 50/50 or 90/100. It depended on information.

I guess in some cases you really need to have specific knowledge. For example for an untrained person like me it would be impossible to say what kind of weapons could have caused the kind of impact or even recognise the weapon on the photo or video footage.

Well, if we sat together for about an hour you would be able to recognise the basic types of weapons. Of course, you will not be able to identify all weapons, especially the ones you do not know, but it is just the beginning of the investigative process.

Do you still need to ask for an expert opinion?

Yes. In some cases we reach out for an expert help. Most of the time, people who help us have military background for example and work for international organisations like the United Nations (UN), or human right organisations such as the Human Rights Watch and Amnesty International. In some cases we ask them to help us identify a weapon. I myself asked for help from an expert when I was doing an investigation into a mosque bombing in Syria where the US was said to have killed many civilians.

How much time did you spend investigating the MH17 crash?

If we speak about all of the Bellingcat team members combined and in hours, I would say weeks, most probably months. After all, it has been three years. I did not work on this investigation, it was my colleagues. But Bellingcat is still doing some work on it. Each time a new piece of information comes in we try to verify it.

So you do not consider it a finished investigation?

No, we do not, though we believe most sources have been exhausted after three years of research.

What do you still want to find out?

It may be additional bits of information. For example, most recently we joined the call of the JIT to identify two persons related to the MH17 incident. It is not an ongoing investigation, but we think some information may still come out. As a big part of this investigation was done by volunteers, some of them might yet find out more.

Do you have any idea about the legal validity of the evidence that you get? Do you think it would stand in the trial?

This is the promise and peril of open source evidence. It is an interesting question, and one I have looked into for my Master research at King’s College London last year. I think we will see quite some development in the coming years related to the use of open source evidence in legal cases. In August 2017, it was the first time the International Criminal Court issued an arrest warrant purely based on footage posted on social media. This is a significant development and I think we are only going to see more of this.

Do you think there are factors at present which could depreciate or undermine publicly available evidence?

Of course. That is why it is important not to rely on one piece of evidence only. We always try to double check and corroborate the evidence we have by independent sources, by eyewitness statements and independent source of information etc. We are trying to do that all the time.

Do you analyse the evidence to check if it was not edited or tampered?

Yes we do it. But there are limits to the extent you can discover such meddling. This is why it is even more important to do cross checks of evidence. In case of MH17, the James Martin Center for Non-Proliferation Studies at the Middlebury Institute of International Studies at Montery used the forensic image analysis software Tungstene to examine various images. Images analysed included photographs of smoke believed to be from the missile launch and images of the Buk missile launcher travelling through separatist territory on July 17th 2014. The analysis found that the images of the Buk missile launcher and what is believed to be smoke from the Buk missile show no signs of digital alteration.

Apart from the MH17 report, Bellingcat also published a report on the shelling of eastern Ukraine from the Russian territory. Was there any specific reason why you got interested in this investigation?

There were two more people from Bellingcat who worked on this. To be clear, I did not work on this investigation. I think it was just a very interesting case to show how one country just shoots at another sovereign country across the border. It is quite something when things like this happen and are being denied, so we wanted to investigate this, and I would say it was a very strong case based on satellite images.

Would you also tackle the Ukrainian army in your investigations?

Of course. I do not care whom I investigate. It gets less public interest but my colleague Aric Toler for example also published information about the actions of the Ukrainian army, how the army used an area very near residential houses for its operations or about the cases when the Ukrainian army violated the Minsk agreement.

We also investigate cases in other regions. For example, in Iraq there are many cases of bombings by the US-led coalition, which were being denied if civilians are killed. We published information about the US army attack on a mosque which was said to be committed by the Russians. In general, I believe it is important to have a strong civil society which can look closely into such cases. We can agree that the strategic communication of western countries can be very good, but for that very reason we should also investigate our own governments and militaries; it is, or should be, part of the checks and balances in a modern liberal democracy.

Christiaan Triebert is an open source analyst with Bellingcat.

Kataryna Pryshchepa is a PhD student at Graduate School for Social Research, Institute of Philosophy and Sociology of the Polish Academy of Sciences.