1 of 16

Giancarlo Fiorella�Senior Investigator, Bellingcat�November 3, 2021

Crowdsourcing for Open Source Investigations

One of the ethical issues that we face in this kind of work is the “spotlight effect”. It’s one thing to geolocate a video shared by a journalist who has 500 thousand followers on Twitter—but what about videos shared by private individuals with 5 followers? What if these private individuals had no intention of their image going viral? In all likelihood, the person who recorded the PS752 missile video didn’t know what he was looking at—and certainly, didn’t know that he was looking at a commercial airliner getting shot by a missile.

When do we shine a spotlight on these videos, of shining a spotlight on them can get the person who recorded or shared them in trouble?

Relatedly, another dilemma that we run into at Bellingcat is that of transparency in sources and methods vs. ensuring the security of the person. Bellingcat is an open source research platform: that means we’re open and transparent about our sources and our methods. But this has limits, obviously. For example, if sharing the source of a video will put that person’s life or well being at risk (because they’ll be arrested by the local authorities, for example). How do we balance these concerns?

2 of 16

Outline

Bellingcat and Crowdsourcing

The January 6^th Capitol Riot

Lessons Learned & Takeaways

The Bellingcat Volunteer Platform

�

3 of 16

Bellingcat and Crowdsourcing

4 of 16

Bellingcat and Crowdsourcing

5 of 16

Bellingcat and Crowdsourcing

6 of 16

The January 6 Capitol Riot

As the Capitol Riot unfolded, we realized that:

1. We were watching a contentious historical event unfold.�

2. The participants were creating a mountain of visual evidence and sharing it on social media in real-time.

We were quickly overwhelmed by the quantity of material coming from the Capitol, so we turned to crowdsourcing to help identify and archive it.

7 of 16

The January 6 Capitol Riot

As the Capitol Riot unfolded, we realized that:

1. We were watching a contentious historical event unfold.�

2. The participants were creating a mountain of visual evidence and sharing it on social media in real-time.

We were quickly overwhelmed by the quantity of material coming from the Capitol, so we turned to crowdsourcing to help identify and archive it.

8 of 16

The January 6 Capitol Riot

9 of 16

The January 6 Capitol Riot

10 of 16

The January 6 Capitol Riot

11 of 16

The January 6 Capitol Riot

12 of 16

The January 6 Capitol Riot

13 of 16

Lessons Learned & Takeaways

There are lots and lots of good people out there with all kinds of skillsets who want to help.

It may be worthwhile for your organization to consider how you might let people help you do your work via crowdsourcing. It doesn’t have to look like what I’ve shown here today.

Think ahead: it’s better to have a tried-and-tested workflow and logistics method than making one up on the spot.

14 of 16

The Bellingcat Volunteer Platform

15 of 16

The Bellingcat Volunteer Platform

16 of 16

Giancarlo Fiorella�Senior Investigator, Bellingcat�November 3, 2021

Crowdsourcing for Open Source Investigations

One of the ethical issues that we face in this kind of work is the “spotlight effect”. It’s one thing to geolocate a video shared by a journalist who has 500 thousand followers on Twitter—but what about videos shared by private individuals with 5 followers? What if these private individuals had no intention of their image going viral? In all likelihood, the person who recorded the PS752 missile video didn’t know what he was looking at—and certainly, didn’t know that he was looking at a commercial airliner getting shot by a missile.

When do we shine a spotlight on these videos, of shining a spotlight on them can get the person who recorded or shared them in trouble?

Relatedly, another dilemma that we run into at Bellingcat is that of transparency in sources and methods vs. ensuring the security of the person. Bellingcat is an open source research platform: that means we’re open and transparent about our sources and our methods. But this has limits, obviously. For example, if sharing the source of a video will put that person’s life or well being at risk (because they’ll be arrested by the local authorities, for example). How do we balance these concerns?