Latest / Stage Talks with Bellingcat / "Building Open Source Research into a Newsroom" with Mike Yambo
Transcript
- 0:04You're listening to a stage talk titled building
- 0:07open source research into a newsroom. To learn
- 0:10how you might embed open source practices into
- 0:13a traditional newsroom, we spoke with Mike Giambo
- 0:16from Kenya's Nation newspaper. Within the talk,
- 0:20he shared invaluable tips and tricks on getting
- 0:23started in OSINT and convincing your peers and
- 0:26employers of the value of the skill set. All
- 0:30of the links mentioned in the talk can be found
- 0:32in this episode's description. This talk was
- 0:35hosted by me, Charlotte Mar, on Thursday the
- 0:3730th of January 2025 in the Bellingcat Discord
- 0:41server. We can, we can hear you very well. Oh
- 1:00yeah, I can see the guys. Yeah, I can see the
- 1:03guys and their bridging. Karavu, welcome! We're
- 1:06all here for the Bellingcat stage talk series.
- 1:09This week we're joined by Mike Yambo from The
- 1:12Nation newspaper in Kenya. Mike is their first
- 1:15and only, I think at this stage, open source
- 1:18investigator on The Nation team and is pioneering
- 1:21a new type of reporting there. I'll let him tell
- 1:25you how well it is going. Before working at The
- 1:28Nation, Mike also had some experience working
- 1:30on investigations with Africa Uncensored. Today
- 1:33we're going to talk about how you might follow
- 1:35Mike's footsteps, setting up open source research
- 1:38in a journalism slash newsroom setting and learn
- 1:41a little more about how Mike communicates his
- 1:43research to wider audiences. We're also going
- 1:46to run through some of Mike's recent investigations,
- 1:49looking at his methods and overall impact. One
- 1:51of which you can already find in our chat, which
- 1:54some of you had a nice read of while you were
- 1:56waiting. As we talk. Please put your questions
- 1:59in the chat. And as I've just said, if you do
- 2:02not want me to read out your Discord username
- 2:04on the podcast recording, please make sure to
- 2:07note that in your question. And I'll make sure
- 2:10to ask those questions as we move into the Q
- 2:12&A section. Okay. Mike, we're ready. Please tell
- 2:16us more. I'm going to mute myself just while
- 2:18you talk a little bit about what you do. All
- 2:24right. Thank you so much, Shelley. I'm truly,
- 2:28truly honored to be here. Thank you for this
- 2:30Pelling Cart invite. It's a great honor. And
- 2:32I must say that I have been a long time for Pelling
- 2:35Cart, just as everyone else. You guys do amazing
- 2:39stuff. My early inspiration for open source investigation
- 2:43was drawn by reading a lot of Pelling Cart's
- 2:46works. And so my name is Mike Yambo. I'm an investigative
- 2:50journalist, a national media group based in Nairobi.
- 2:54My work focuses on combining open source investigative
- 2:57techniques with traditional journalism. I'm here
- 3:01to share my experiences on not only pioneering
- 3:04open source investigations in our newsroom, but
- 3:07also share a journey of gradually building a
- 3:10new way of investigative storytelling vaccination.
- 3:12I also have a huge love for storytelling, which
- 3:16is why I'd like to share this personal experience
- 3:19through a story from how I entered the newsroom.
- 3:22to working on actual open source investigations.
- 3:26Through this story, which I'll try to compress
- 3:29into the few minutes I've got, I believe you
- 3:32understand how we build an open source research
- 3:35perspective nation, and why OSINT has become
- 3:38crucial to our investigative journalism. So when
- 3:42I was hired at Nation, an open source investigative,
- 3:45as an open source researcher, OSINT was a gray
- 3:48area in the newsroom. The traditional culture
- 3:51of reporting was dominant in our reports, and
- 3:54new techniques like fact -checking and using
- 3:56tools weren't very common. So I was hired as
- 4:00an investigative researcher who will focus on
- 4:02infusing open source investigative techniques
- 4:04with traditional reporting. Our newsroom was
- 4:08at the time heavily shifting towards public interest
- 4:11stories focused on investigative journalism,
- 4:14and open source journalism was a way of bringing
- 4:17forward our investigative storytelling. It was
- 4:20an immense challenge of not only debunking news
- 4:23and disinformation, but also facing now open
- 4:25-source -led investigations. I think we all understand
- 4:29that all cities that released really new concepts
- 4:31in the newsrooms, especially the mainstream media,
- 4:34who are kind of now gradually trying to catch
- 4:36up. And this was also my fate. I also realized
- 4:40that most of our editors were struggling to understand
- 4:42this kind of storytelling. You know, only a few
- 4:45did. I remember my initial stories ended up re
- 4:49-edited to suit the traditional way of storytelling.
- 4:52Some ended up being published late, way beyond
- 4:55the headlines, while some ended up not seeing
- 4:58the light of the day. So when I say stories were
- 5:02re -edited to fit the traditional way of storytelling,
- 5:05what this is what I simply mean. Some key components
- 5:08of posting stories like, you know, archiving
- 5:10links, hyperlinks, or even visualizations ended
- 5:14up being removed. And that was very frustrating,
- 5:16obviously, because that meant that I ended up
- 5:20with an incoherent piece of work. Something I
- 5:24found also particularly hard was educating my
- 5:30seniors as to what awesome pieces of work needed
- 5:33to look like. If I just imagine that you would
- 5:35be trying to educate an old journalist who has
- 5:37been here, who has been in Tech TM for probably
- 5:3920 plus years. In most cases, that will look
- 5:42like disrespect to them. And so I only explain
- 5:44my approaches to the ones who kind of bother
- 5:47to ask me, and some militias who are particularly
- 5:50worried that it took too long to deliver the
- 5:53story. Remember, this is like a newsroom environment.
- 5:56Things are moving so fast, you know, and catching
- 5:59up with some of these things while using awesome
- 6:02methodologies might be particularly hard. But
- 6:06I think even as the militias are worried about,
- 6:08you know, our turnaround times, Uh, I probably
- 6:12think they were right to a certain degree because
- 6:14often stories are told from an evidence point
- 6:16of view. And the fact checking and clarification
- 6:19is usually quite a lengthy process most of the
- 6:22time. But generally it was, it was a micro job
- 6:25in the beginning to be honest. There was generally
- 6:27a slow culture to publish the fact checks I work
- 6:30on, um, with a speed at which other pieces got
- 6:33published. But I kind of had somehow braced myself
- 6:37for it. because they understood it would be hard,
- 6:40based on the news culture that existed way before
- 6:43my arrival. But interestingly, slowly by slowly,
- 6:49let me say, probably after a year, story after
- 6:53story, I began noticing gradual shifts. As some
- 6:56editors then call it, it began, it started becoming
- 6:59curious to understand what Osint was and what
- 7:03it took to produce such kind of reports. So instead
- 7:07of struggling with so many obvious explanations,
- 7:10which is often boring, I believe what I did was
- 7:14share my articles with the ones who were particularly
- 7:16interested. In that regard, I believe that it
- 7:20was easy to make them truly conceptualize the
- 7:22open source methodologies to build an even more
- 7:26deeper sense of clarity on, you know, to make
- 7:30people truly understand and appreciate this kind
- 7:32of storytelling. What I did was simply collaborate
- 7:36on stories with other reporters, incorporating
- 7:38individual investigative elements into their
- 7:41stories. So from this proposition, we have produced
- 7:44major public interest journalism that are deploying
- 7:48open source techniques. We have debunked false
- 7:52claims from Kenya's Directorate of Criminal Investigations,
- 7:57just Kenya's principally investigative agency
- 7:59under the National Police Service. And subsequently,
- 8:03our news reports resulted in a statement of public
- 8:07apology from the DCI, which is a very big win
- 8:11when it comes to public interest journalism.
- 8:14We have also used remote sensing tools to tell
- 8:17other stories of how a religious cult that led
- 8:21to the deaths of hundreds of Kenyans secretly
- 8:24established itself inside a forest for two years.
- 8:28Most people did not really realize that this
- 8:30was going on. And this was, this popularly became
- 8:34known as the Shakahola Massacre, where we used
- 8:37Sentinel Hub and we used Google Earth to kind
- 8:40of, you know, establish and determine when did
- 8:44the, you know, the cows start building their
- 8:48homes inside this forest in the name of prayers.
- 8:52And most of them ended up dying as a result of
- 8:55starvation. You know how cows can turn out to
- 8:58be. And we consistently also monitored protests,
- 9:03and in the process, analyzed police brutality
- 9:05incidents using these officers' techniques. We
- 9:08have also analyzed Twitter networks, particularly
- 9:12a Twitter network revealing how hashtags were
- 9:14deployed to spread news and disinformation claims
- 9:17against Kenyan doctors who went for 90 days'
- 9:22protests due to a poor working environment. And
- 9:27discovered that this got probably government
- 9:30and anti -drug government bloggers who are trying
- 9:33to play a particular agenda to spread some sort
- 9:37of false claim against the doctors who are worried
- 9:41and very concerned about their working environment.
- 9:45We have also told stories about the current war
- 9:49raging in Sudan between the RSS and the Sudan
- 9:52armed forces. And this story, we largely pivoted
- 9:57our data from the Sudan Weakness Project under
- 10:02the Center for Information Resilience. Toward
- 10:052025, that's the last, 2024, sorry, that's the
- 10:09last year, my amazing colleague at the research
- 10:11desk, Jacqueline Mosharia, did a brilliant work
- 10:15using open source information to track cases
- 10:17and establish a database on femicide statistics
- 10:20in Kenya. So on story after story, our newsroom
- 10:25nation has produced investigations that have
- 10:28really relied on Austin. I know that that's quite
- 10:33a long story about how we began this and where
- 10:37we are so far. But before I conclude, I think
- 10:41I'd like to drop a piece of advice, like general
- 10:45advice to someone or a newsroom that intends
- 10:47to venture into OSINT. One thing I've come to
- 10:51realize or kind of understand is that venturing
- 10:53into open source intelligence is an individual.
- 10:57or newsroom requires some sort of elements like
- 11:01strategic planning, skill development, and even
- 11:04some ethical considerations before venturing
- 11:07into it. So the first, I believe, is of understanding
- 11:11all things is truly get from the standard statistics
- 11:14of all things if you want to do it for a large
- 11:16newsroom or even from an individual capacity.
- 11:20All things involves collecting, analyzing, and
- 11:22very fine publicly available information. And
- 11:26the best way to be good in all things, I believe,
- 11:28is to know how to find data, verify data, and
- 11:32how to analyze data. The second thing that I
- 11:38also kind of would like to pass to people who
- 11:41are interested in developing this kind of storytelling
- 11:44is to invest in training and skill development.
- 11:49Media houses can invest in training their journalists
- 11:51on the use of verification techniques in their
- 11:54reports. I have conducted trainings for journalists
- 11:57as well as attend many trainings myself as well.
- 12:01And as an individual journalist, I identify an
- 12:06awesome community or personality that does these
- 12:09kinds of works. On a personal note, I taught
- 12:13myself how to do this kind of journalism to be
- 12:16honest. And I did this by closing the YouTube
- 12:20tutorials of Benjamin Streep. who works for,
- 12:23who currently works for the Center for Employment
- 12:25and Resilience. He's been so very instrumental
- 12:27in my, you know, laying the foundations of my
- 12:30investigative, operational journalism. Um, and
- 12:34if you'd like to learn the fundamentals of awesome
- 12:37reporting, I highly recommend you watch the past
- 12:39videos. They are very, very rich, and they kind
- 12:41of help you as a journalist nail the, uh, the
- 12:45why, the where, and the when. He kind of uses
- 12:48tools to explain this, you know, basic concept.
- 12:51concepts of a journalistic piece. I think I've
- 12:56also found Bellingcat's website having tons of
- 13:00free guides and resources, which can be very
- 13:03fundamental, especially if you need to develop
- 13:05your skills. Another website that offers free
- 13:08guides, particularly on video verification and
- 13:11AI is witness .org. vast resources on how to,
- 13:19you know, navigate the AI space, which is now
- 13:21becoming a very elusive and a very big subject
- 13:25when it comes to the walls of investigations.
- 13:28And also generally AI, it's to kind of help broaden
- 13:32your mental scope when you need to understand
- 13:36this kind of stuff around AI. Something else
- 13:40for a newsroom or an individual who intends to
- 13:43like put up an authentic research test, is to
- 13:47learn the basic tools like, you know, Google
- 13:50Jogging, how do you resign your searches online?
- 13:52How do you go and find that particular piece
- 13:56of information that you're trying to find online?
- 13:59Google Jogging simply saves you time by being
- 14:03very particular on what you specifically want
- 14:09to find online. Another basic thing that one
- 14:14can... uh, prob - probably investing is, uh,
- 14:17knowing how to scan websites, knowing how to
- 14:20establish the validity of websites, for instance.
- 14:23Uh, and, and I know who is local tool is a very,
- 14:26is a very good tool that was like way back machine.
- 14:29Uh, there are basic things like, you know, performing
- 14:32reverse image searches. I believe these are very
- 14:35basic elements that one or a particular newsroom
- 14:39that intends to deploy this kind of closed source
- 14:41methodologies can... easily invest on and produce
- 14:45a story that can be, you know, the turnaround
- 14:47time is very fast. Something else as you build
- 14:52your awesome toolkit is to learn the basic methods
- 14:56of social media analysis. There are tools out
- 15:00there. Most of them basically are paid, but some
- 15:04of them are paid that can help you analyze social
- 15:07media. For instance, I know X. You can use Mel's
- 15:11Water Explore, uh, to, you know, analyze the
- 15:14hashtags trends. Uh, you also analyze the sentiment
- 15:17analysis I've used before in my reports. Uh,
- 15:21the famous report about the storming of Caliament,
- 15:23we began by, you know, analyzing the hashtags,
- 15:26seeing what people were saying online before
- 15:29transferring their page, uh, into the streets
- 15:31of Nairobi and other, other streets of, of, of
- 15:35Kenya. These also, uh, the most basic elements
- 15:39of posting, which I believe most people should
- 15:42really, really invest in. Because I think that's,
- 15:45that is what kind of gave me the confidence around
- 15:48all things, with the geolocation and satellite
- 15:50analysis. Um, and one master, master tool to
- 15:56help you perform this is the Google Earth Pro,
- 15:58uh, which has tons and tons of features that
- 16:01can help you to, you know. geo -rotate and also
- 16:04establish the locations of different things or
- 16:07situations that you're trying to find out. Sentinel
- 16:10hub is also a very good one when it comes to
- 16:15establishing changes over time in a particular
- 16:17place. Geolation and satellite analysis, I think,
- 16:21is one of the most common practices around osintric
- 16:24now. And it's one thing that can easily instill
- 16:27the confidence in you to be able to perform other
- 16:30aspects of OSINT. Something else is to develop
- 16:34a verification culture in your newsroom. This
- 16:38is something about encouraging a mindset of being
- 16:41skeptical about things that you see online. Just
- 16:46because something is online, it doesn't mean
- 16:49it's true. And so if you are able to perform
- 16:53a Google reverse image search, for instance,
- 16:56and do a whole report about it, or just a very
- 16:59brief report about it, It's going to be very
- 17:02vital in establishing that foundational level
- 17:05of, of, of audience, um, in which most process
- 17:08I like, I like referring to as part checking.
- 17:12Also is heavily reliant on part checking. And
- 17:14so if you are able to perform small part checks
- 17:17over time, then you'll be able to perform even,
- 17:20um, complex part checks and even, uh, perform
- 17:24complex person's investigations, the deep dive
- 17:27over time. So you don't have to like start so
- 17:30big. You can start with a small factor and over
- 17:33time you'll truly understand and learn the basic
- 17:37and the different tenets that, you know, are
- 17:41tied to OSINT reporting. It's also very important
- 17:46to, even as you look for it, to start out your
- 17:50OSINT research depth in your newsroom or even
- 17:53intend to do this on a particular or like on
- 17:56an individual level. Let's go to ensure your
- 18:00digital security and also look at the ethical
- 18:03considerations around this field of open -source
- 18:07research. It's very vital to be aware of the
- 18:10legal risk. So just understand what the law points
- 18:14out concerning data in your own country and follow
- 18:17ethical guidelines to avoid or violation of privacy,
- 18:22which might end up in court cases and you may
- 18:26probably end up with lawsuits or something which
- 18:29you do not want to enter into. The second last
- 18:34thing about integrating officers reporting into
- 18:37your reports or like into your newsroom is to
- 18:41apply awesome techniques in your projects. I
- 18:43think I've mentioned this, please a little bit.
- 18:47And it doesn't have to be mega reports at first,
- 18:50just simple fast checks as you slowly gain momentum
- 18:52over time. Other way to integrate open source
- 18:58into your newsroom is probably also having a
- 19:02small dedicated team working around awesome staff
- 19:06or even just hiring an awesome specialist within
- 19:10the newsroom. And this will prove very significant.
- 19:13It can help in, I believe, two ways. It can save
- 19:16you time verifying information if your newsroom
- 19:20is so quick and fast, and also it will help you
- 19:23at the core. add the complementing elements of
- 19:25a story, for instance, visualizing data using
- 19:28data wrapper. So some of these tools and techniques
- 19:31can help you, you know, make your stories more
- 19:33rich and even more compelling to your audiences.
- 19:37And finally, liberate collaboration and open
- 19:39source communities. Some of the most successful
- 19:42audience investigations that I've seen, for instance,
- 19:45have one underlying thing in them, and that is
- 19:47collaboration. Collaborate with partners and
- 19:50researchers to improve reporting. As you already
- 19:54see, Atonation, for instance, we have active
- 19:57partnerships with other research organizations
- 19:58like Coach for Africa. CFA basically has access
- 20:02to some awesome tools that we do not have at
- 20:04Atonation. Therefore, we are able to, you know,
- 20:08have a sense of partnerships while doing some
- 20:11stories. I hope this was helpful when, you know,
- 20:15looking on how to set an awesome tool of this
- 20:18in your newsroom. Thanks so much for listening
- 20:21and I'll hand it over back to you, Charlie, for
- 20:24the Q &A session. Mike Nakashikuru, thank you
- 20:29so much for that amazing brief. You went through
- 20:34so many tools just then that somebody in the
- 20:36chat said, this is like the most informative
- 20:39talk I've ever attended, has so much practical
- 20:41information for newcomers like me. So thank you
- 20:44so much for running through so many tips. And
- 20:48before I get to audience questions, I just wanted
- 20:50to ask you, what kind of sparked your passion
- 20:53in open source research? What was the moment
- 20:56where you decided to invest time into this skillset?
- 21:02Interesting. Yeah, so I just made a completed
- 21:05compass. And at that point in time, the BBC had
- 21:09released something on Cameroon, a documentary
- 21:12called There Has to Be a Bad Healing, which I
- 21:14believe is very popular with office structures.
- 21:18And I was very curious about how this was done,
- 21:22that level of masterpiece, and the fact that
- 21:25they were able to produce a very impactful piece
- 21:28of journalism. So I got really curious, and I
- 21:31wanted to know how they did it. And over time,
- 21:36basically, I'm a curious person. I tend to read
- 21:38a lot, especially when I discover new stuff.
- 21:42So just navigating the whole aspect of Google
- 21:46Maps, Uh, I started, I came across Benjamin Street's
- 21:51YouTube channel. Um, and he was there teaching
- 21:54so many things about awesome. But at that point
- 21:56in time, I didn't know that he was the, he was
- 21:59part of the team that was behind the anatomy
- 22:02of a killing. Um, and so I found these tutorials
- 22:06very, very educative. And I think with the foundational
- 22:10level of fact checking that I got at African
- 22:13Censored. I got to build up over time by watching
- 22:17Ben's videos on YouTube. And Ben also cited Bellingcat
- 22:22in most of these examples. So it was what, it
- 22:25was some kind of a ripple effect. And over time,
- 22:29I wanted to draw this nepotism, so I ended up
- 22:32practicing. And slowly by slowly, that's how
- 22:36I got into art scenes. And coincidentally, I
- 22:41got a job a year later. Um, well, funny how I
- 22:44was learning all these things, uh, without being
- 22:47actively engaged in the journalism space. So
- 22:50when, um, an interesting opportunity to be, uh,
- 22:53an investigative research admission came up in
- 22:56that power was like able to, you know, get in
- 22:59and nail and execute some of these things that
- 23:02had learned over time. So, uh, just to answer
- 23:06you precisely that anatomy of a killing documentary
- 23:08sparked my, um, my enthusiasm for open source
- 23:12reporting. Amazing. I just shared the link to
- 23:15that in the chat. I believe some of our team
- 23:19helped on that, but I might be wrong. If anybody
- 23:22from the team then, back then, can comment. I
- 23:27remember there's some relation to us, but anyway,
- 23:31it's an amazing investigation if you haven't
- 23:33had it. a look at it. So please do check that
- 23:37out. And I also linked, while you were talking
- 23:40earlier, Bender Brown's YouTube channel as well.
- 23:42So please look at that and Code for Africa as
- 23:45well, while you're going through it. We have
- 23:47quite a few audience questions coming in. Following
- 23:50off your recent reporting on the storming of
- 23:55parliament, someone's asked, how do you, when
- 23:58fact checking government sources, how do you
- 24:01ensure the safety of your newsroom? when you
- 24:03go up against the government, for example? That's
- 24:08a particularly interesting question. So for the
- 24:10newsroom itself, we have this set of ethical
- 24:14guidelines we have to operate around. Especially,
- 24:19you know, now when it comes to forming of the
- 24:21parliament for instance, there was a way we are
- 24:25supposed to handle the coverage. because it was
- 24:31a very hostile environment at that point in time,
- 24:34if you can't remember. And I have some Kenyan
- 24:37audiences here, who on ground reporting became
- 24:40particularly hard for journalists themselves.
- 24:43So, one of the things that we were taught to,
- 24:49you know, or kind of something that was kind
- 24:52of highlighted before we got off to the field
- 24:54at that point in time during the protests is
- 24:57to make sure that you try and provide the accurate
- 25:03information that is on the ground. Because there
- 25:07was a lot of anger at that point in time and
- 25:09any misreporting or any misrepresentation on
- 25:14what was actually going on on the grounds would
- 25:18have led to, you know, this public backlash or
- 25:21outrage. So we were very keen in terms of, you
- 25:24know, our reporting, trying to ensure the accuracy
- 25:27and also trying to not really, um, not really
- 25:32miss out on some of the facts and the fundamental,
- 25:34uh, aspects of reporting. Uh, just as a general
- 25:39to keep, you know, the general media house image
- 25:43out there so well. We, we didn't really, we didn't
- 25:46really want to, you know, of a report or kind
- 25:48of, you know, over an alive situation, you know,
- 25:52being the session that means sensational about
- 25:56our reports. Um. That was one of the approaches
- 26:02that we used while covering the process, just
- 26:05being very factual and being very accurate. But
- 26:09something else that you can look out for, especially
- 26:13when you are an on -scene researcher, it's being
- 26:17able to prevent the doubts and to be able to
- 26:20prevent a lot of questioning around the works
- 26:25that you do. is to be able to show the methodology,
- 26:29the approach in which you do your reports. They
- 26:33need to be as transparent as possible, but also
- 26:36that's not enough because coming from a journalistic
- 26:40background and even at Nation, we try to blend
- 26:43these methodologies to the actual stories or
- 26:50other storytelling aspects. kind of merging what
- 26:54our findings, the research findings, and also
- 26:58the reports and also the, the reports and, you
- 27:02know, the things that the interviews that we
- 27:05carry out, marry them together and find a very
- 27:08accurate account for our reports. And that is
- 27:13what we also did during the process. So it really
- 27:16helped us go a long way in terms of being accurate
- 27:19and, you know. Being able to tell the truth out
- 27:22there to what was really happening, not only
- 27:24Kenya, but also just clearly to the entire globe.
- 27:27It's really got a lot of, you know, global attention.
- 27:31Absolutely. As we always say at Bellingcat, stick
- 27:34to the facts and always communicate that open
- 27:39source research can only get you so far as well.
- 27:41And that you're not coming with all of the answers,
- 27:44but with as much that open source can tell you.
- 27:47Um, it's just a starting point for investigations
- 27:49as we like to get out. Um, so we have a few more
- 27:53questions and I want to make sure that we get
- 27:55to them before the end of the hour. So Sarah's
- 27:58asked, hi Mike, I have to use Facebook a lot
- 28:01to dig into the social media of the community
- 28:04I'm curious about. And I wondered if you had
- 28:07any tips for researching on Facebook. I can't
- 28:10do something like create a list on Twitter and
- 28:13it's extremely time consuming. Do you have any
- 28:15research tips for Facebook? have I researched
- 28:20on Facebook recently? Not really. Yeah, but I
- 28:27mean, Facebook is a very interesting area and
- 28:32it's one of those social media platforms that
- 28:36I've been really interested into. And then also
- 28:41considering that some of the tools were brought
- 28:45down and also the most recent developments we
- 28:47are checking. was discontinued. It's something
- 28:53that we are considering, that we are really,
- 28:55really upscaling as a new group. You know, how
- 28:57to tell more stories, especially regarding the
- 29:01digital threats around Facebook, because we have
- 29:04so much happening there in terms of scamming
- 29:07and also in terms of, you know, potential mis
- 29:10and disinformation. So it's an area that I've
- 29:13not really... put on so much weight in terms
- 29:17of advising newsletter. It's something that,
- 29:21you know, I'd really be interested in learning
- 29:23about also as well. Even as I continue doing
- 29:27further investigations, much of my investigations
- 29:30in recent times have heavily revolved around,
- 29:34you know, the dark social media sites like the
- 29:38Telegram. And I've also been heavily doing open
- 29:41source visual investigations. But Facebook is
- 29:44a great area and it's something that I think
- 29:49we work on so much moving forward in our newsroom,
- 29:54for sure. Yeah. Everybody in the chat was like,
- 29:59your size said it all. I think we're all feeling
- 30:01this kind of disappointment in how easy, what
- 30:06a lack of... control we have over kind of Facebook
- 30:09research. You mentioned Telegram. So I've just
- 30:12popped into the chat, one of Bellingcat's tools.
- 30:16Yeah. So we have a tool called the Telegram phone
- 30:19number checker. So I've just popped that into
- 30:21the chat for anybody who's regularly researching
- 30:24on Telegram. It's a really useful tool. So check
- 30:26that out on our GitHub. And thank you for going
- 30:30through all that Mike though. In terms of, you
- 30:32mentioned earlier that your newsroom was a little
- 30:36bit reluctant to take on open source at the beginning.
- 30:40What kind of changed your editor's minds? What
- 30:43kind of galvanized the change? Because you mentioned
- 30:47to me before the talk that your team was actually
- 30:50looking to grow. So what has kind of convinced
- 30:53your editors or your newsroom that open source
- 30:55is worth the investment? Yeah, for sure. One
- 31:02thing, one thing is producing stuff and, you
- 31:05know, getting your stuff out there. Because I
- 31:09remember first it was so hard, you know, doing
- 31:12his reports and doing this project and see them,
- 31:15you know, timely to, you know, the traditional
- 31:18news that he calls. But it's very important now,
- 31:23for my case, then... What I had at that point
- 31:27in time was producing so many reports and presenting
- 31:30to my editors and, you know, also being able
- 31:33to assess the value of whatever I was bringing
- 31:35in the newsroom at that point in time, because
- 31:38we are trying to introduce the new thing altogether.
- 31:42Not so many people are familiar about it. And
- 31:46so you need to have a few reports to show some
- 31:49of the editors about the impact of your work.
- 31:52Just remember, it's easier to sell an idea when
- 31:54you have the actual work itself. And something
- 31:59else that is very important, especially to change
- 32:01the mindset and shift people's ideas and how
- 32:07they view OSINT is to emphasize on the impact
- 32:12of OSINT -driven investigations, such as exposing
- 32:16work crimes, for instance. In Africa, we have
- 32:18so much that is happening. around Sudan and,
- 32:22you know, Congo right now. You can produce such
- 32:26kind of, you know, war crime stories to truly
- 32:29make them understand the methodology, the process,
- 32:32and also probably the impact. Verifying this
- 32:36information or tracking illicit financial flows
- 32:39demonstrates how async complements traditional
- 32:42reporting. This, you know, most editors, whenever,
- 32:46especially in the traditional newsroom setups,
- 32:49as long as you can save the resources, save the
- 32:52money, and somehow improve your storage capabilities,
- 32:59then you see the interest gravitating towards
- 33:02your work. So basically, it's about having reports,
- 33:07impactful reports, that shows some of the editors
- 33:10the impact of your work, and also trying to emphasize
- 33:14the impact of passing prevention investigations.
- 33:19to the editor so that they can see the real value
- 33:22and the real need to do this kind of journalism.
- 33:26In that sense, I believe it will be easier to
- 33:28sell the awesome idea to your newsroom bosses.
- 33:32Thank you for that. We actually had a question
- 33:35earlier from Alicia, who was asking for some
- 33:38more information about how to find your Sudan
- 33:41work. So I've just linked in the chat, one of
- 33:44your pieces called, Look, the fire is inside
- 33:46your houses. Where are you? Sudan's war within.
- 33:52But if there's any other particular links that
- 33:54you want to highlight, please pop them in the
- 33:56chat if you can, Mike, because I think there's
- 33:58quite a want to have a read of some of your Sudan
- 34:02coverage as well. Oh yeah, sorry. So that Sudan
- 34:06investigation was just kind of inspired by the
- 34:11need to, you know, report so practically on Sudan
- 34:14because so much was going on there. Um, I also
- 34:18saw Falling Country something, NYT did something.
- 34:21Um, but it was also kind of a way of adding up,
- 34:25you know, getting the voices out there to the
- 34:28world to truly understand the, uh, the situation
- 34:30instead of that. It, it took us about three years
- 34:33to kind of nail it down. Uh, but I believe it's
- 34:36one of those pieces that had some sort of impact,
- 34:40especially, and now that you are... kind of highlighting
- 34:42the damage and the losses that Sudan underwent
- 34:46between 2023 and 2024. It was a very powerful,
- 34:52awesome piece. It's one that I keep on responding
- 34:55to whenever probably I'm training or I'm giving
- 34:59examples to some of my works. Yeah, there's so
- 35:03much in Africa. There's so much in Ethiopia currently.
- 35:06So, yeah, but I see the link to my Sudan was,
- 35:09it's a very classic example of how to switch
- 35:13your locations and locations. That's basically
- 35:16what you did. Absolutely. And there's so, so
- 35:19many stories on the continent at the moment,
- 35:22unfortunately, to do with conflict as well. As
- 35:24you mentioned, we've been publishing on Burkina
- 35:27Faso as well. Just had a story come out yesterday
- 35:31by some of the members in the audience on the
- 35:35Democratic Republic of Congo as well. So thank
- 35:38you so much for sharing that and to you Sudan
- 35:41experience in terms of your coverage. And you
- 35:44can see that from that article and also from
- 35:47your coverage of the stormy parliament and you,
- 35:50you do a lot with visualizations and something
- 35:53that I really connected with when I looked at
- 35:55your work was how. well, you communicate kind
- 35:59of the findings that you've come up with in your
- 36:02research. What are some of the factors that you
- 36:06consider when thinking about visualizations?
- 36:08What are the key factors you consider when thinking
- 36:11about how you're going to visually tell the story?
- 36:17First of all, it's to communicate the impact
- 36:20of that incident we are trying to find out. I
- 36:24believe through pictures and you know, uh, videos,
- 36:28it's very, very, um, easier to, you know, sell,
- 36:32sell a story or like kind of communicate the
- 36:35impact of something that you're trying to communicate
- 36:37out there. But, you know, doing visual investigations
- 36:42and, you know, all that stuff around, you know,
- 36:45geolocations and location, you have to consider,
- 36:48uh, if that piece of media you're trying to analyze
- 36:52answers, um, Because now I'm saying this from
- 36:56a journalistic point of view, if it answers the
- 37:00four W's on the H in a journalistic piece, but
- 37:04also the cases, you'll want to look into the
- 37:08who, the why, and the how. So if a particular
- 37:12piece of video, for instance, that is associated
- 37:15with a Kenyan protest, you want to understand,
- 37:18maybe you're trying to analyze a police brutality
- 37:21incident. you want to understand where that image
- 37:25was probably taken, when it was taken and who
- 37:28was behind it. Those are some of the factors
- 37:31that I consider myself in my works and I think
- 37:35probably other people do as well. And also on
- 37:38top of that, it can kind of help you to weave
- 37:40through your story and also add on top the element
- 37:45of storytelling aspect and the voices, you know,
- 37:48the sources voices on top of it. So, Basically,
- 37:54nailing and understanding the context and trying
- 37:58to answer those questions particularly well and
- 38:01having them come up so clearly will inform the
- 38:04decision to have this piece of visual evidence
- 38:08as a journalistic piece of work. So basically,
- 38:12it's analyzing the whole, the way, the how, and
- 38:15generally piecing it up together, adding on top
- 38:18the sources' voices. you know, kind of paints
- 38:22the context and the events very clearly. Yeah,
- 38:28because for us, it tells stories from our journalism
- 38:31prior to Rio. I think I've read much of Bellingham's
- 38:35works. It's usually mostly based on research.
- 38:41But then now for us at Nation, we try to add
- 38:43the journalistic elements on top of it to, you
- 38:47know, kind of communicate, you know. the wider
- 38:50problem, what is the human story in that data
- 38:54point or in that particular event that you're
- 38:56trying to answer with the four W's and a H. So
- 39:00that is our approach when it comes to visual
- 39:04storytelling. And it's incredibly powerful. And
- 39:08I love how many different modes and mediums you
- 39:12use to tell those stories as well. Even referring
- 39:14back again to the storming the parliament piece,
- 39:17you have 3D models in there. I've seen cases
- 39:20where you've used peak Pfizer to showcase the
- 39:23mountain ranges that you found. I think it's
- 39:26really amazing that lots of different methods
- 39:29that you use in telling your stories. Yeah, basically
- 39:33just to, you know, add the juice to these kind
- 39:37of stories, make them even more relevant to people.
- 39:40Because I believe the whole essence of doing
- 39:43journalism is to be able to communicate your
- 39:46story out there, make people understand and truly
- 39:49engage them. Because there's no whole point of
- 39:54showcasing all these brilliant skills that without
- 39:57really addressing the real, real concern that
- 40:01needs to be addressed. So as much as, you know,
- 40:06I've borrowed all these spelling cards types
- 40:08of storytelling, in as much as we do heavy coding
- 40:12to our stories, we try to kind of help drive
- 40:17some sort of impact. And that's why we particularly
- 40:21add the human voices on top of our stories. I
- 40:25think that is a very powerful way to, you know,
- 40:28put your work out there as a journalist. Not
- 40:30only not focusing on the methods, You know, for
- 40:33instance, you mentioned the 3D models. They are
- 40:36very, you know, cool ways of telling a story,
- 40:39but we use them to make the people remember that
- 40:44story even further. Yeah. The innovation has
- 40:48to fuel understanding. We have one other question
- 40:51from the audience. Guilfoyle asks, and I think
- 40:54this is based on some of your comments earlier
- 40:57about ethically and doing investigations, particularly
- 41:01when you're starting out and also thinking about
- 41:04how you stay secure as well. Guilfoyle said,
- 41:07how do you handle operational security among
- 41:10your whole team? I imagine you get several phishing,
- 41:14like email phishing attempts given your line
- 41:16of work. Is that something that you have had
- 41:19to face, particularly since starting in open
- 41:21source research where you're kind of digging
- 41:23into large data sets? Have you had a lot of online
- 41:27threats or online attacks at all that you've
- 41:29had to deal with within your team? Yeah, interesting.
- 41:34I mean, at the media house, we were attacked
- 41:36with a series of hashtags some point last year
- 41:39as a result of our coverage that we had all these
- 41:42forms of means and disinformation spread against
- 41:46the mission media group. And it was politically
- 41:50driven, basically. But now how can you, you know,
- 41:53stay safe while doing all this kind of stuff.
- 41:57Myself, I've not really gone so deep, especially
- 42:01to the aspects of, you know, maybe trying to,
- 42:06you know, track someone. But generally, I tend
- 42:12to think this whole concept of, you know, being
- 42:15secure online, probably like even having the
- 42:18virtual boxes, sometimes it tends to be really,
- 42:21really talking much about, but an essential thing
- 42:26generally, and this is something that I picked
- 42:30up over time and I also read so much about it.
- 42:33It's really good to use secure communication
- 42:38channels as you try and, you know, probably reach
- 42:41your sources or try to contact people who maybe
- 42:45you believe are in some sort of danger. And there
- 42:50are tons of them out there, you know, secure
- 42:52communication channels. We have session, we have
- 42:55proton mail, we have saving. I believe as open
- 42:58source researchers, I believe we have an ideal
- 43:00space. But if you don't, there are tons out there
- 43:04that, you know, maybe you can't read until you
- 43:07understand how to, you know, use them. Something
- 43:12else is just basically avoid logging into your
- 43:15personal accounts while maybe trying to track
- 43:17down an individual. In 253, we did an investigation
- 43:21about, it was a PEP, a person of interest, like
- 43:25a politically exposed personality who was connected
- 43:29to the Al -Shabaab terrorism in East Africa.
- 43:35And so, as I was researching, there was so much,
- 43:39tons and tons of information about this person
- 43:40on Facebook. At that point in time, uh, what
- 43:45I did was establish, uh, another second account
- 43:48to help me kind of, you know, obscure myself.
- 43:51Because this guy basically would post stories
- 43:54on a daily basis. And if you probably do a mistake
- 43:59of, you know, going and feeling the story using
- 44:02your, uh, real account on that you used to connect
- 44:05to your loved one or one that is so personal
- 44:08to you. they can be able to, you know, see your
- 44:11digital footprints or kind of see you as one
- 44:13of the viewers to that particular story. So having
- 44:16a second account or a banner account is very,
- 44:19very powerful. Also protecting yourself by knowing
- 44:25the legal risk associated with, you know, Austin's
- 44:28because some Austin techniques may violate the
- 44:30hyper -fiction laws. VPNs are there. I know they've
- 44:34been talked so much about. I use it myself most
- 44:37of the time. As a company, we have a policy.
- 44:41Our emails are well protected in -house. We usually
- 44:45change our passwords after every three weeks
- 44:47or one month, I believe. So it kind of helps
- 44:50us secure our communication channels. So whenever
- 44:53you're trying to establish a connection with
- 44:56a source, who probably maybe is a whistleblower,
- 45:01it kind of helps, it kind of limits the threats
- 45:05of hacking or phishing. or all that stuff that
- 45:09involves security concerns. Yeah, so the changing
- 45:13of passwords, but that is done for us automatically
- 45:16in our media house. And also all those other
- 45:19things that are mentioned, they're very particular,
- 45:22particularly important while, you know, doing
- 45:25this online repurchase. And it helps quite a
- 45:31long way in terms of even keeping our own...
- 45:34your loved ones that you are, people who are
- 45:36so close to you. Amazing. I know we started a
- 45:39little bit late, so I want to ask one last question,
- 45:42if that's okay, if you have time. Yeah, absolutely.
- 45:45We can have all this time. Amazing. The question
- 45:48I wanted to ask, and it's kind of a summary of
- 45:50what you've been speaking about. I wanted to
- 45:52understand from your perspective, what has been
- 45:55the... Best parts and the most difficult parts
- 45:59of working with open sources for breaking news
- 46:02stories, like fast moving stories. What are the
- 46:05best and the most difficult aspects of working
- 46:09with open source in that context? Interesting
- 46:12question. What else in the country competes with
- 46:17breaking news stories? Yeah, because breaking
- 46:20news stories physically are meant to like, you
- 46:22know, move quickly. The story needs to be updated
- 46:27and people move on after that. But the best thing,
- 46:32if you ask me, there's an investigation I did
- 46:34in 2023 around July, and my turn around time
- 46:40was particularly faster. So we had this protest
- 46:43in Kenya. It was like an anti -government protest
- 46:46led by the opposition leaders. And we had a lot
- 46:50of police brutality meted out against opposition.
- 46:54Strongholds in Kenya. And there was this particular
- 46:59video that really went viral online towards the
- 47:04end of those demonstrations on a particular evening.
- 47:08So there were a group of police, 16 police officers,
- 47:11and unleashing brutality towards a single person
- 47:15who was lying there helplessly. So when this
- 47:20photo was shared to me by my editors, Um, it
- 47:24was around 10 p .m. If my memory serves me right.
- 47:29Um, and by midnight had like kind of hacked a
- 47:33geolocation for that particular piece of, of
- 47:37video. And we were able to establish where it
- 47:39happened, the when, and all those, all those
- 47:43basic questions one asked themselves while maybe
- 47:46analyzing a piece of video. Um, the following
- 47:50day, uh, that particular Geolocation was used
- 47:53by human rights organizations. I remember Amnesty
- 47:57International reaching out to me with a Twitter
- 47:59handout. They used that to, you know, go and
- 48:03provide help to the affected families in the
- 48:07aftermath of those protests because there was
- 48:09so much violence in the opposition strongholds.
- 48:12So I did, that was one of my successful stories,
- 48:16which like I had a quick turnaround. And it helped
- 48:20not only in our newsroom, but it also helped
- 48:24the public interest journalism that people are
- 48:27speaking about. Because you see human rights
- 48:30organizations, civil society organizations using
- 48:33a piece of work to help these families. That
- 48:37in itself is a very humbling moment for you and
- 48:40your career as a journalist. Failures, we only
- 48:48have failures, especially when, you know, but
- 48:50we were doing open source research or we, maybe
- 48:54when we are getting started. Keeping up with
- 48:56the new cycle can be so hard. Um, one thing that
- 49:01we've not really, really, um, paint on a personal
- 49:06level up not really nails so well, um, is, you
- 49:10know, applying all those tools. We have so many
- 49:13tools. and which are quite expensive to, you
- 49:16know, have access to. So some of the stories
- 49:20that, potential stories that I identify online,
- 49:24like coordinated inauthentic behaviors online,
- 49:28or particularly on X, I tend to miss out on some
- 49:31of those stories on the day they happen. But
- 49:36then through collaborations, I mentioned we collaborate
- 49:39with Code for Africa. We're usually able to reach
- 49:42out later to them and they provide us with the
- 49:45necessary data that can help us tell that story
- 49:48after a day or two or even a week. So we do not
- 49:55really have access to all the tools, you know,
- 49:57some of them are really expensive, but we have
- 50:00access to majority of the tools basically. So
- 50:04that usually makes me feel somehow bad about,
- 50:08you know, meeting out and, you know, missing
- 50:11out on retorting some of these aspects. But usually
- 50:14they make up for that reporting on that incident
- 50:17way later, like the day to reporting aspects
- 50:19to it. So we don't do it as a breaking news story
- 50:22or a regular news story. We do it as a day to
- 50:26story, a piece of story. Yeah. And that's how
- 50:30I've been able to navigate my success and failure
- 50:34and it's been impactful generally. Such good
- 50:39advice, because I know that's something that
- 50:41a lot of people struggle with is kind of the
- 50:44fast moving nature of news and how to cover it
- 50:48with open source as quickly as possible. I mean,
- 50:50at Bellingcat, we still struggle with that sometimes
- 50:53in terms of mobilization, particularly when there's
- 50:56multiple news stories a day. So thank you so
- 51:00much for that advice, Mike. As onto Sana, it's
- 51:03been amazing to have you. In the beginning, you
- 51:06said that you were a fan of Bellingcat. I think
- 51:08you've created a fan of Mike's, to be honest,
- 51:12in the audience, myself included. Thank you so
- 51:16much for this talk. It's been amazing. And some
- 51:20of our mods have put in the chat. Please hand
- 51:22it over. Thank you so much, Ali and Bellingcat
- 51:24for having me. Some of our mods are put in the
- 51:27chat. Please hang around the Discord server because
- 51:29there's quite a lot of people who want to ask
- 51:31you questions in the server. So please do hang
- 51:33around afterwards as well and just have a little
- 51:37browse. Thank you so much for this talk and thank
- 51:40you to everybody who's joined in the audience
- 51:42as well today. Thank you for listening to the
- 51:45stage talk. If you'd like to catch a stage tour
- 51:48live where you can ask the guest questions, join
- 51:52the Bellingcat Discord server by visiting www
- 51:55.discord .gg slash Bellingcat. The music you've
- 52:01heard is titled Dawn by Newer Self and is courtesy
- 52:05of Artlist.