Placeholder Image

字幕列表 影片播放

  • "Good morning.

  • My name is Malachy Browne.

  • I'm a senior producer on the Visual Investigations team

  • here at the 'Times.'

  • Thanks for tuning in and sending questions in advance.

  • We've got dozens of questions from viewers and readers

  • of the 'Times,' which we're going to go through.

  • First of all, we're going to introduce the team."

  • "I'm Haley Willis."

  • "Hey.

  • I'm Evan Hill.

  • I'm a visual journalist on the team."

  • "I'm Barbara Marcolini."

  • "Christoph Koettl."

  • "And Christiaan Triebert."

  • "And one of the first questions that we got

  • is what is open-source investigation?

  • So Christiaan, have a go."

  • "Yeah, so basically, open-source investigation

  • is it's reporting, but using any kind

  • of openly-available source.

  • So think of a Facebook post, or a tweet, a YouTube video,

  • or just a database-- anything you can find online, openly,

  • and for free.

  • So this is the opposite of closed source.

  • Like if a source tells me something or tells

  • my colleague something, you will

  • need to believe that source.

  • But the strength of open-source reporting

  • is that anyone with an internet connection

  • and a laptop can access that same source

  • and can take the same steps for verification.

  • So that's what we do here--

  • a lot of open-source investigation."

  • "And a lot of the process, if there is an event

  • that we're investigating, it involves, as Christiaan said,

  • collecting as much evidence as we possibly can,

  • you know, from the open web.

  • There's so much documentary visual evidence out there now

  • that allows us to get to the truth of an event

  • to break it down and really analyze it moment by moment.

  • And so that could be video satellite imagery,

  • a timestamp from a tweet, but it also

  • involves turning those open sources who

  • are witnesses into primary sources,

  • and finding secure ways to get in touch with them,

  • and find out more about what they saw,

  • very often get the raw imagery as we did in Syria.

  • Do you want to talk about that a little bit?"

  • "Yeah, I think that was an interesting one

  • because that combined sort of traditional reporting

  • methods with the open-source stuff

  • that I've been learning as a new member of the team.

  • But for that story for Syria, we

  • needed to get a wealth of material."

  • "Do you want to explain what the series was about just

  • for a second?"

  • "Yeah, so it was a series of stories about Russian bombing

  • of hospitals in Syria and other civilian targets that

  • came out over the course of last year.

  • And we needed to get a bunch of material

  • from the ground-- videos and photos

  • from Syrian journalists.

  • And so that was basically who we

  • can find through WhatsApp networks,

  • through activist networks, people

  • we can reach out to and expand our sources on the ground

  • speaking in Arabic a lot of the time."

  • "And the key was using the material

  • to identify the minute that a strike happened."

  • "Yeah, and that included metadata analysis, so

  • going into the actual files, explaining

  • to the sources on the ground the best way

  • to send the files to us so that we can extract

  • the information from the file to confirm,

  • oh, yes, we heard the airstrike occurred

  • at 5:00 PM.

  • We'll the file tells us, yes, it did occur at 5:00 PM."

  • "And by establishing the very minute of those attacks,

  • then we obtained access to thousands

  • of intercepted recordings of Russian pilots

  • carrying out their missions.

  • And so by understanding the time that something happened

  • on the ground, we could then examine

  • what was going on in the skies at that time, and then,

  • you know, basically apportion blame

  • to specific pilots for attacks on hospitals, on an IDP camp,

  • a busy commercial street in Syria.

  • So the evidence that witnesses are collecting

  • and that we're managing to verify through our processes

  • here can be really valuable to the journalism.

  • How did you get your start in OSINT?

  • OSINT is Open-Source Intelligence.

  • It's a term that lots of people

  • use to describe this work.

  • We call it visual investigations

  • or open-source investigations.

  • Choose your term, but how did you get your start

  • in it, Barbara?"

  • "I actually started working at Storyful.

  • Storyful is an agency that verifies

  • content on social media.

  • So that's how I started, just working at this agency

  • and learning from their experience

  • there, which by the way, Malachy

  • used to work for them."

  • "I came from Storyful too.

  • That was where I got my start too.

  • Haley?"

  • "I got my start working with a student collective

  • actually at the University of California, Berkeley

  • at our Human Rights Center.

  • We had a lab that did this kind of work

  • for human rights legal cases and for advocacy groups.

  • We had a partnership with Amnesty International

  • and their Digital Verification Corps.

  • So that's kind of how I started this work is using

  • the same open-source techniques

  • that everyone's been talking about,

  • and we applied that to kind of human rights issues."

  • "Christoph?"

  • "I got my start at Amnesty International in 2007,

  • so quite a while ago.

  • And back then, I think when I started at Amnesty,

  • it's the first time that a human rights group started

  • using satellite imagery.

  • So that was my-- well, that is my speciality.

  • We did a lot of work around satellite imagery

  • for a few years, and then around 2011,

  • it became just extremely important with the Arab

  • Spring to verify YouTube videos, photos,

  • and similar materials.

  • So we had to teach ourselves, basically, how to verify that

  • and how to integrate that into human rights reporting."

  • "Yourself?"

  • "So I was an aspiring journalist as a student,

  • so I tried to do photo reports in Iraq, and Syria,

  • and Ukraine, but I felt like I'm not

  • the best writer there is in the world

  • and I'm also not the best photographer.

  • So I felt like what am I doing?

  • What am I contributing to what is already out there, right?

  • Anyone with an internet connection

  • can find stories that are better

  • written, have better photos.

  • And I was intrigued by a guy called Eliot Higgins, who

  • started Bellingcat, an open-source investigation

  • group.

  • And I just started tweeting.

  • Literally on Twitter, I started tweeting out

  • my findings because in Iraq, I saw airstrikes

  • in the distance, but I didn't know much about it

  • except for I could say, well, an airstrike happened.

  • But by using that same kind of satellite imagery Christoph

  • was talking about, I could figure out

  • when did the airstrike or what did it

  • target, who was in control of the village.

  • And I started tweeting out those findings and got

  • involved in 2015-- so wow, five years ago--

  • with this group called Bellingcat.

  • And I've been doing that ever since before I joined

  • the 'Times' last year."

  • "A great question to get started on the practices

  • from Natalie in Toronto and Sena in Tehran--

  • how do you verify a video?

  • There's lots of different ways you can verify video,

  • but maybe we'll cut to a show and tell.

  • We have one from recently when the Ukrainian airliner

  • crashed down outside Iran.

  • Iran was denying that it was shot down.

  • There were reports at the time that it was allegedly

  • due to technical malfunction.

  • And over the course of a few days,

  • we had started mapping out the evidence related

  • to the downing of that airliner.

  • It killed over 170 people on board.

  • Maybe we'll just walk through that.

  • You've got a slide show that I'll cut to here.

  • Sorry, not that one.

  • It's on Keynote.

  • Bear with me for a second.

  • Yeah, there we go."

  • "Right, so let's just start with it right?"

  • "Yeah."

  • "So PS752, it basically started for us

  • with a Slack message from a colleague,

  • who was like, whoa, a plane has been crashed in Iran.

  • And we were all on high alert.

  • It was the evening here in New York.

  • Basically, the whole team was still in the office

  • because Iran had just launched those ballistic missiles

  • onto military targets in Iraq.

  • So we were all on high alert.

  • And then this message comes in and we're like,

  • O.K., whoa, let's investigate it.

  • So one of the first things we're doing with this

  • is open-source reporting.

  • We explained it in a lot of words,

  • but it's an actual example.

  • This is a plane crash.

  • Now, any commercial airliner nowadays is being tracked.

  • And anyone with an internet connection

  • can track those flights through websites

  • like Flightradar24, RadarBox24.

  • And here, we see the actual flight path

  • of the plane that allegedly crashed near Tehran.

  • And we can see how it's here on the--

  • it's starting to take off.

  • It's departing the International Airport

  • near Tehran, the capital of Iran.

  • And we can see how it flies away.

  • And we can also see the flight path is not that long.

  • It basically stops right here.

  • Now, the great thing is that we can also

  • download this flight path for free

  • and put it in a program that's called Google Earth.

  • And Google Earth is just basically

  • Google Maps on steroids.

  • It's, as you can see here, this is the same flight path

  • that we just saw on the website.

  • It's now on Google Earth and you can see it's also in 3D.

  • The height is also visible.

  • And we can see exactly where the last transponder sign

  • of that flight was spotted.

  • That's the first start for us, just

  • to have basic information about the incident

  • and have a sense of time and space.

  • Now, it started to get morning already in Iran.

  • And before people started uploading photos

  • from the crash site, people saw that plane going down.

  • And you can see it here.

  • There are different videos.

  • Here, we see the plane, but we couldn't verify those videos

  • at first.

  • But what we did is we matched them up with this explosion

  • to be sure, hey, that they are showing the same incident.

  • So these videos seem to be from the same plane crash,

  • so it's a first step.

  • We hadn't verified the location of this,

  • but the crash sites were--

  • sorry-- anyone nowadays has--

  • usually, a lot of people have a mobile phone, right?

  • And Tehran is a massive city.

  • So if something like this happened what you just

  • saw in those videos is that people start filming

  • something when it happens.

  • I mean, think of yourself.

  • Imagine something big is happening.

  • And here, we can see a video from the actual crash site

  • when the morning has broken already.

  • And one method we use, usually as a first step for verifying

  • a video to come back to the question,

  • is a process we call geolocation.

  • And geolocation is determining where a photo or a video

  • has been taken based on visual clues in the video.

  • Well, let's have a look.

  • We see a lot of debris.