WEBVTT

00:00:04.659 --> 00:00:08.000
You're listening to a stage talk titled How We

00:00:08.000 --> 00:00:12.759
Got Here, The Rise, Backlash and Feature of Counterdisinformation.

00:00:13.339 --> 00:00:15.759
In this episode, we're joined by Bellingcat founder

00:00:15.759 --> 00:00:18.940
Elliot Higgins and Cornell University professor

00:00:18.940 --> 00:00:22.320
Dr. Claire Wardle. Together, they unpacked the

00:00:22.320 --> 00:00:25.359
historical context of online fact -checking and

00:00:25.359 --> 00:00:28.160
where open source investigations really began

00:00:28.160 --> 00:00:31.359
to shine. You can find links to all the resources

00:00:31.359 --> 00:00:33.619
mentioned in the talk in the podcast description.

00:00:34.079 --> 00:00:36.219
This talk was hosted by me, Charlotte Ma, on

00:00:36.219 --> 00:00:39.719
Thursday the 29th of January, 2026, in the Villain

00:00:39.719 --> 00:00:52.119
Cap Discord server. Hello, beautiful people.

00:00:52.420 --> 00:00:55.960
We're back in a new year with a new set of talks

00:00:55.960 --> 00:00:59.530
to kick off our 2026 series. We're discussing

00:00:59.530 --> 00:01:02.409
the rise, fall and future of counter disinformation

00:01:02.409 --> 00:01:04.909
and fact -checking efforts. Something central

00:01:04.909 --> 00:01:07.489
to the work we do at Bellingcat and important

00:01:07.489 --> 00:01:10.269
to the balance of our society. Someone who's

00:01:10.269 --> 00:01:12.450
been a leader in the study of the way we receive

00:01:12.450 --> 00:01:14.989
and respond to information for decades is Dr

00:01:14.989 --> 00:01:17.349
Clare Wardle. Claire is a co -founder of the

00:01:17.349 --> 00:01:19.709
former First Draft News, a non -profit providing

00:01:19.709 --> 00:01:21.989
practical and ethical guidance in how to find,

00:01:22.129 --> 00:01:24.650
verify and publish content sourced from the social

00:01:24.650 --> 00:01:27.129
web. She's currently an associate professor at

00:01:27.129 --> 00:01:30.250
Cornell specializing in information ecosystems

00:01:30.250 --> 00:01:32.349
with a focus on effective communication in the

00:01:32.349 --> 00:01:35.069
age of AI and conspiracy. If you're familiar

00:01:35.069 --> 00:01:37.269
with Bellingcat's work, our other speaker needs

00:01:37.269 --> 00:01:39.609
no introduction. We're joined by Bellingcat founder

00:01:39.609 --> 00:01:42.000
Elliot Higgins. Elliot has been working with

00:01:42.000 --> 00:01:43.939
the University of Nottingham on a new framework

00:01:43.939 --> 00:01:46.540
demonstrating the importance of epistemic security

00:01:46.540 --> 00:01:49.200
for a healthy democracy. Today they'll be breaking

00:01:49.200 --> 00:01:51.579
down the emergence and journey of counter disinformation

00:01:51.579 --> 00:01:54.439
strategies and answering your questions on what

00:01:54.439 --> 00:01:57.519
our future looks like as people are now fed information

00:01:57.519 --> 00:02:00.299
according to political preferences and life experiences.

00:02:00.879 --> 00:02:03.319
Whilst we talk you can place your questions in

00:02:03.319 --> 00:02:05.280
the chat accessible on the right hand corner

00:02:05.280 --> 00:02:07.719
of your screen but please remember that this

00:02:07.719 --> 00:02:10.870
is being audio recorded for the podcast So if

00:02:10.870 --> 00:02:13.229
you don't want me to mention your username, please

00:02:13.229 --> 00:02:16.569
add that as a note in your question. Okay. Elliot,

00:02:16.750 --> 00:02:20.889
Claire, over to you. Oh, yeah. Thanks for coming

00:02:20.889 --> 00:02:24.789
to speak to us, Claire. I think I first met you

00:02:24.789 --> 00:02:29.129
probably 20, my memory's terrible. So 2011, 2012,

00:02:29.330 --> 00:02:34.050
the very early days of this whole kind of, I

00:02:34.050 --> 00:02:37.050
guess, movement you could call it. This, correct

00:02:37.050 --> 00:02:38.550
me if I'm wrong, this started for you when you

00:02:38.550 --> 00:02:40.669
were in Storyful and maybe you can tell us a

00:02:40.669 --> 00:02:42.789
bit about what Storyful was and how you kind

00:02:42.789 --> 00:02:47.120
of came into this area. Yeah, so actually it

00:02:47.120 --> 00:02:49.000
goes a little bit earlier than that. I was at

00:02:49.000 --> 00:02:51.599
Cardiff University as a professor and wrote a

00:02:51.599 --> 00:02:54.199
report for the BBC about user -generated content,

00:02:54.580 --> 00:02:56.159
because in 2008 they were like, we're getting

00:02:56.159 --> 00:02:57.639
this stuff from the audience and we don't know

00:02:57.639 --> 00:02:59.699
what's true. So I was like, oh, interesting.

00:03:00.039 --> 00:03:01.680
And so I wrote the report thinking that only

00:03:01.680 --> 00:03:03.919
my mom was going to read it. And one person from

00:03:03.919 --> 00:03:05.479
the BBC read it and said, oh, would you come

00:03:05.479 --> 00:03:08.259
for six months and help us implement your recommendations,

00:03:08.400 --> 00:03:11.110
which never happens as an academic. And while

00:03:11.110 --> 00:03:13.330
I was there, many of you may remember this, but

00:03:13.330 --> 00:03:15.569
that plane landed in the Hudson River in January

00:03:15.569 --> 00:03:18.409
2009. And the head of BBC News Gavin called me

00:03:18.409 --> 00:03:20.750
and said, you're the Twitter lady. None of my

00:03:20.750 --> 00:03:22.349
journalists are on Twitter. We got scooped by

00:03:22.349 --> 00:03:24.770
CNN because we couldn't find that image. We couldn't

00:03:24.770 --> 00:03:26.310
verify it and we didn't know whether we could

00:03:26.310 --> 00:03:28.310
use it ethically or legally. Can you develop

00:03:28.310 --> 00:03:31.050
a training course for the BBC on how do we verify

00:03:31.050 --> 00:03:33.750
and use this kind of content? So I just was like,

00:03:33.830 --> 00:03:36.330
had this niche area of like, how do you verify

00:03:36.330 --> 00:03:39.500
what's on this? social web and then did that

00:03:39.500 --> 00:03:41.659
worked for a couple of years independently and

00:03:41.659 --> 00:03:42.939
traveled the world training journalists. I mean,

00:03:42.939 --> 00:03:45.060
I was so lucky and it was an extraordinary time

00:03:45.060 --> 00:03:47.759
to go to Al Jazeera and to go to ABC Australia

00:03:47.759 --> 00:03:51.099
and go to BBC Africa and like many countries

00:03:51.099 --> 00:03:53.659
in Africa to train them in how do you tell whether

00:03:53.659 --> 00:03:55.330
or not the audience is telling the truth. But

00:03:55.330 --> 00:03:57.990
then from that in 2012, I moved to Storyful when

00:03:57.990 --> 00:04:00.969
it was still very small, with an incredibly small

00:04:00.969 --> 00:04:04.050
group of people trying to figure out the Syrian

00:04:04.050 --> 00:04:05.710
war. And at the time, I didn't even know your

00:04:05.710 --> 00:04:07.370
real name, right? Like you were just, they were

00:04:07.370 --> 00:04:09.830
like, this is Guy, he's in Leicester, he's really

00:04:09.830 --> 00:04:12.349
good at munitions stuff. And you were like, you

00:04:12.349 --> 00:04:14.189
know, Moses Brown. And we didn't, we never met,

00:04:14.289 --> 00:04:16.410
we just knew you. And so there was like this

00:04:16.410 --> 00:04:19.389
crew at Storyful, this small crew at BBC, anti

00:04:19.389 --> 00:04:21.569
-Carvin, NPR and you. And we were just trying

00:04:21.569 --> 00:04:26.129
to figure this stuff out. tell people about that

00:04:26.129 --> 00:04:27.850
kind of period. It was such a small group of

00:04:27.850 --> 00:04:30.230
people who are kind of figuring out exactly even

00:04:30.230 --> 00:04:33.170
what words to use. Open source investigation,

00:04:33.170 --> 00:04:35.670
it wasn't even a term we were using. Someone

00:04:35.670 --> 00:04:39.230
started using Austin and that kind of stuck and

00:04:39.230 --> 00:04:41.009
that's kind of plaguing us a bit now, I think,

00:04:41.110 --> 00:04:43.670
unfortunately. A good way to spot hoses as well.

00:04:44.350 --> 00:04:48.620
But also, Words like geolocation, I can't remember

00:04:48.620 --> 00:04:50.860
where that word came from, but it kind of then

00:04:50.860 --> 00:04:53.899
become just the word we use to talk about what

00:04:53.899 --> 00:04:56.240
we now call the process of geolocation. But for

00:04:56.240 --> 00:04:57.660
me, at the first time, I spot the difference

00:04:57.660 --> 00:05:01.420
with Google Maps. Yeah. Well, because at story

00:05:01.420 --> 00:05:03.480
for the time, there was Malachy Brown, who then

00:05:03.480 --> 00:05:05.660
went on and many people probably know, works

00:05:05.660 --> 00:05:07.240
for the New York Times. And I remember at the

00:05:07.240 --> 00:05:09.500
time he wrote a blog post for us, which was like

00:05:09.500 --> 00:05:12.339
the steps to verify a video. And I think I said

00:05:12.339 --> 00:05:13.439
to him at the time, I was like, I think this

00:05:13.439 --> 00:05:15.519
might one day go into a museum, Malachy. Like

00:05:15.519 --> 00:05:18.610
it was this crazy attempt by one individual to

00:05:18.610 --> 00:05:20.629
try and lay out the steps that he was doing every

00:05:20.629 --> 00:05:23.290
day. And, and I think lots of that was everybody

00:05:23.290 --> 00:05:24.730
was helping each other to be like, well, what

00:05:24.730 --> 00:05:26.589
are you doing there? And, you know, and I remember

00:05:26.589 --> 00:05:29.290
the first time that we tried to analyze an explosion,

00:05:29.370 --> 00:05:31.509
a distance of an explosion. And we were like,

00:05:31.509 --> 00:05:33.290
Oh, has anybody done this before? Like, how do

00:05:33.290 --> 00:05:35.189
you figure that out forensically? So yeah, it's

00:05:35.189 --> 00:05:36.730
kind of extraordinary what was happening at the

00:05:36.730 --> 00:05:39.389
time. And I don't think we realized how important

00:05:39.389 --> 00:05:42.220
that time was. Yeah, I think that's kind of your

00:05:42.220 --> 00:05:44.079
experience when you do this work because everything's

00:05:44.079 --> 00:05:46.939
so new. You're saying how like measuring the

00:05:46.939 --> 00:05:48.300
explosion. That's something I remember doing

00:05:48.300 --> 00:05:50.379
in a completely separate case, measuring the

00:05:50.379 --> 00:05:53.399
range of a rocket launcher and based on how far

00:05:53.399 --> 00:05:55.399
away the noise of its rocket explosion was from

00:05:55.399 --> 00:06:01.800
the origin. So yeah, I mean, and I guess Syria

00:06:01.800 --> 00:06:04.959
was the kind of dominant topic in 2011 and 2000.

00:06:05.339 --> 00:06:07.740
probably till about 2013, I would say, especially

00:06:07.740 --> 00:06:10.639
with the sarin attacks. But eventually you moved

00:06:10.639 --> 00:06:16.300
on to another organization from Storyful. Yes.

00:06:16.800 --> 00:06:18.660
Well, I just want to remind, and I had just another

00:06:18.660 --> 00:06:20.939
moment. I remember one of the early chemical

00:06:20.939 --> 00:06:23.480
weapons attacks where we all figured out that

00:06:23.480 --> 00:06:25.879
the different platforms had different timestamps,

00:06:25.879 --> 00:06:28.060
right? Because Putin had said that it had happened

00:06:28.060 --> 00:06:29.850
yesterday and we were like, what? And it was

00:06:29.850 --> 00:06:31.709
because we hadn't figured out that YouTube was

00:06:31.709 --> 00:06:33.610
Pacific Standard Time. And then we're like, but

00:06:33.610 --> 00:06:34.889
that's different to Twitter, because Twitter

00:06:34.889 --> 00:06:37.529
depends on where your settings are versus all

00:06:37.529 --> 00:06:39.629
of this stuff now that we take for granted. But

00:06:39.629 --> 00:06:41.709
it's incredible to remember those kind of moments

00:06:41.709 --> 00:06:44.110
that were so defining. But yeah, after StoryFile,

00:06:44.170 --> 00:06:47.009
I went back and did a fellowship for a year of

00:06:47.009 --> 00:06:49.730
updating that user -generated content research.

00:06:49.750 --> 00:06:51.709
I did that at the TAL Center of Digital Journalism.

00:06:51.990 --> 00:06:54.310
And then around that time, founded a nonprofit

00:06:54.310 --> 00:06:57.560
called Eyewitness Media Hub. which was a small

00:06:57.560 --> 00:06:59.980
group of us trying to figure out like what are

00:06:59.980 --> 00:07:02.060
the ethical uses of eyewitness media? And I have

00:07:02.060 --> 00:07:03.920
to say this week there's been some incredible

00:07:03.920 --> 00:07:07.399
pieces written about eyewitness videos from Minnesota.

00:07:07.800 --> 00:07:09.720
And obviously that's what we've all been looking

00:07:09.720 --> 00:07:12.139
at for a long time, but it's kind of amazing

00:07:12.139 --> 00:07:14.899
to think back to, there were real questions around

00:07:14.899 --> 00:07:17.300
that time, like the Charlie Hebdo shooting. I

00:07:17.300 --> 00:07:19.040
don't, people may not remember this, but there

00:07:19.040 --> 00:07:22.860
was a guy in his apartment and he filmed the

00:07:22.920 --> 00:07:26.439
killing by the security forces and then uploaded

00:07:26.439 --> 00:07:29.199
to Facebook, not thinking. Somebody else scraped

00:07:29.199 --> 00:07:31.360
it from Facebook, uploaded it to YouTube and

00:07:31.360 --> 00:07:34.259
monetized it. And all the newsrooms were using

00:07:34.259 --> 00:07:36.459
the YouTube version, not realizing that they

00:07:36.459 --> 00:07:38.579
should check the Providence, right? And so again,

00:07:38.680 --> 00:07:40.680
there was such uncertainty in those early days

00:07:40.680 --> 00:07:42.879
of like, what were guidelines for newsrooms?

00:07:43.180 --> 00:07:45.879
What were the best? ethical practices. And I

00:07:45.879 --> 00:07:47.660
think we're seeing some of the conversations

00:07:47.660 --> 00:07:50.019
around Minnesota. I'm like, Oh, actually, there's

00:07:50.019 --> 00:07:51.980
a whole host of earlier stuff, because I got

00:07:51.980 --> 00:07:54.740
interested in this with the 2005 777 bombings.

00:07:55.040 --> 00:07:58.500
which was the first time the BBC ever led a news

00:07:58.500 --> 00:08:00.579
broadcast with anything from a camera that wasn't

00:08:00.579 --> 00:08:03.579
BBC owned. Somebody was down in the tube tunnels

00:08:03.579 --> 00:08:06.319
and had not an iPhone, because it was 2005, a

00:08:06.319 --> 00:08:08.100
grainy photo from inside. And it was the first

00:08:08.100 --> 00:08:09.680
time the BBC had led that, right? So when we

00:08:09.680 --> 00:08:14.319
think about the tsunami in 2004, 7 -7 bombings,

00:08:14.639 --> 00:08:16.060
blah, blah, blah, it's kind of a fascinating

00:08:16.060 --> 00:08:17.959
story of the kind of evolution of eyewitness

00:08:17.959 --> 00:08:20.720
media. Yeah, and I think as well, I've been thinking

00:08:20.720 --> 00:08:23.519
about the past 20 years in terms of how our whole

00:08:23.519 --> 00:08:27.139
information ecosystem has changed. And in parallel

00:08:27.139 --> 00:08:29.199
to all the kind of really dark stuff that's been

00:08:29.199 --> 00:08:31.620
happening and obviously that we battle against

00:08:31.620 --> 00:08:34.220
every day now. There was also this growth of

00:08:34.220 --> 00:08:36.340
this parallel community that were truth seekers,

00:08:36.440 --> 00:08:38.000
but trying to figure out actually how do you

00:08:38.000 --> 00:08:40.460
establish the truth? I remember, I think the

00:08:40.460 --> 00:08:42.820
big challenge for a lot of newspapers and media

00:08:42.820 --> 00:08:46.299
organizations were citing people like the gay

00:08:46.299 --> 00:08:48.179
girl in Damascus blog, which turned out to be

00:08:48.179 --> 00:08:50.000
a white guy in America. And that really shook

00:08:50.000 --> 00:08:51.799
their faith in their ability to trust that they

00:08:51.799 --> 00:08:54.100
can go online. It seems silly now, you shouldn't

00:08:54.100 --> 00:08:56.480
trust stuff on the internet, but back then all

00:08:56.480 --> 00:08:58.419
kinds of stuff was being reported on just because

00:08:58.419 --> 00:09:02.129
it became It seemed legitimate. I do remember

00:09:02.129 --> 00:09:04.549
one really bizarre one. There's something awful

00:09:04.549 --> 00:09:06.610
for a music called Martin Random who did a series

00:09:06.610 --> 00:09:09.549
of posts about the Bush administration talking

00:09:09.549 --> 00:09:12.210
about Donald Rumsfeld having to wear iced underwear

00:09:12.210 --> 00:09:14.129
and all kinds of ridiculous things. It got picked

00:09:14.129 --> 00:09:16.190
up by the Daily Cast and went a bit semi -viral

00:09:16.190 --> 00:09:19.529
because of it. Just those basic things now that

00:09:19.529 --> 00:09:21.570
would be completely ridiculous to do just weren't

00:09:21.570 --> 00:09:24.309
established within journalism and the blogosphere

00:09:24.309 --> 00:09:26.950
also had a big influence that suddenly we were

00:09:26.950 --> 00:09:30.009
getting these alternative voices. people didn't

00:09:30.009 --> 00:09:33.070
know who was kind of the respectable voices very

00:09:33.070 --> 00:09:35.950
early on. So you got that kind of weirdness emerging

00:09:35.950 --> 00:09:39.129
from that. Yeah. And of course it wasn't Syria,

00:09:39.450 --> 00:09:41.929
but Hurricane Sandy, which was the end of October

00:09:41.929 --> 00:09:45.549
2012, was the first time I remember seeing like

00:09:45.549 --> 00:09:47.389
all sorts of fake, you know, people like, oh

00:09:47.389 --> 00:09:49.690
my God, look at this, like the Statue of Liberty.

00:09:49.690 --> 00:09:52.070
And it was from like some disaster movie or,

00:09:52.070 --> 00:09:54.009
you know, really obvious, or, you know, there

00:09:54.009 --> 00:09:58.129
were users that were talking about tubes, subways

00:09:58.129 --> 00:10:00.009
being flooded and they weren't. And I think that

00:10:00.009 --> 00:10:01.990
when I had this moment of like, hang on, there

00:10:01.990 --> 00:10:04.470
was so much more misinformation that up until

00:10:04.470 --> 00:10:06.389
that point I was training newsrooms saying like,

00:10:06.470 --> 00:10:08.669
well, once in a while you might get hoaxed. And

00:10:08.669 --> 00:10:10.330
that was the moment I was like, oh no, hang on,

00:10:10.370 --> 00:10:12.509
this is a completely different level and how

00:10:12.509 --> 00:10:14.750
in real time a newsroom is meant to verify this

00:10:14.750 --> 00:10:17.490
kind of content. When do you think was the kind

00:10:17.490 --> 00:10:19.850
of moment you really saw that sea change? What

00:10:19.850 --> 00:10:23.289
kind of year would you say that was? So actually,

00:10:23.429 --> 00:10:25.129
I think even though it sounds silly, but I think

00:10:25.129 --> 00:10:26.570
it was Hurricane Sandy because I think there

00:10:26.570 --> 00:10:29.009
was something about Syria, which was like there

00:10:29.009 --> 00:10:31.129
was nobody else in, like there were no journalists

00:10:31.129 --> 00:10:33.149
allowed into Syria around that time. So we were

00:10:33.149 --> 00:10:35.690
only able to rely on YouTube. And so a lot of

00:10:35.690 --> 00:10:38.269
it was like, can we verify that that bomb did

00:10:38.269 --> 00:10:41.110
land on that day? Even though that footage is

00:10:41.110 --> 00:10:43.090
probably from around this time, we were trying

00:10:43.090 --> 00:10:45.110
to be really specific. I think Hurricane Sandy

00:10:45.110 --> 00:10:47.610
was like, oh, some people just want to like cause

00:10:47.610 --> 00:10:49.970
like create content and call shit just because

00:10:49.970 --> 00:10:52.830
they can. Like that for me was this moment of

00:10:52.830 --> 00:10:54.750
like, oh, something different is happening here.

00:10:55.029 --> 00:10:56.490
Like, which I think up until the stuff that we

00:10:56.490 --> 00:10:58.029
were looking at Syria didn't feel like that at

00:10:58.029 --> 00:11:02.490
all. I started feeling around probably 2013 is

00:11:02.490 --> 00:11:04.529
when the social media companies were engaging

00:11:04.529 --> 00:11:06.250
a lot more with this. I remember, of course,

00:11:06.330 --> 00:11:09.090
there's the Google Conflict and Connected Worlds

00:11:09.090 --> 00:11:12.289
Summit in late 2013, where it was the first time

00:11:12.289 --> 00:11:14.429
I kind of personally had done a major event,

00:11:14.610 --> 00:11:17.070
which was terrifying for me. And there are lots

00:11:17.070 --> 00:11:18.950
of people there from like the Syrian kind of

00:11:18.950 --> 00:11:21.240
community and journalists and It was the first

00:11:21.240 --> 00:11:23.679
time I'd seen people talk to each other in that

00:11:23.679 --> 00:11:26.120
way and people caring about that because before

00:11:26.120 --> 00:11:27.820
that it was just me and a few people on the internet,

00:11:28.100 --> 00:11:31.360
including yourself, doing this. Where did you

00:11:31.360 --> 00:11:33.399
start feeling this was something that was being

00:11:33.399 --> 00:11:36.120
taken more seriously by the social media companies?

00:11:39.299 --> 00:11:42.840
So it's a good point. I mean, in 2014 I worked

00:11:42.840 --> 00:11:45.580
for a year at UNHCR as like head of social media

00:11:45.580 --> 00:11:47.899
and at the time we had real problems with smugglers

00:11:47.899 --> 00:11:51.179
basically creating content of like beautiful

00:11:51.179 --> 00:11:54.740
pictures of yachts using UNHCR logo and that

00:11:54.740 --> 00:11:56.799
was 2014 and we were trying to contact Facebook

00:11:56.799 --> 00:11:59.429
and others and they were like freedom of expression,

00:11:59.649 --> 00:12:01.230
everybody has a right to pretend to be a smuggler,

00:12:01.370 --> 00:12:03.750
maybe smugglers are good. These kind of crazy

00:12:03.750 --> 00:12:06.210
conversations with these Silicon Valley bros.

00:12:07.750 --> 00:12:10.389
So I think around that time, they weren't, newsrooms

00:12:10.389 --> 00:12:12.389
were. When I was doing interviews with newsrooms

00:12:12.389 --> 00:12:15.429
in 2013, 2014, they were like, we are terrified

00:12:15.429 --> 00:12:17.429
of getting it wrong. And I'd be like, so are

00:12:17.429 --> 00:12:19.669
you building up these skills in -house? And they

00:12:19.669 --> 00:12:22.440
were like, no, we just use Reuters and AP. BBC

00:12:22.440 --> 00:12:24.899
was building it in -house, but many, many newsrooms

00:12:24.899 --> 00:12:26.259
were like, we don't know where to start with

00:12:26.259 --> 00:12:28.460
this kind of content. I think the social media

00:12:28.460 --> 00:12:30.820
company as well, I think the point that you're

00:12:30.820 --> 00:12:35.340
probably trying to get to is in, it was 2015,

00:12:35.759 --> 00:12:38.899
basically Google invited a small group of us

00:12:38.899 --> 00:12:41.580
to a New York meeting. They didn't pay for our

00:12:41.580 --> 00:12:44.100
flights. But we all were like, oh, because at

00:12:44.100 --> 00:12:45.759
the time we were like, oh, Google, exciting.

00:12:45.879 --> 00:12:47.279
It's like Willy Wonka's factory, right? And we

00:12:47.279 --> 00:12:50.600
all, you were there. There was people from Storyful,

00:12:50.779 --> 00:12:53.500
Eyewitness Media Hub, Medan. There were different

00:12:53.500 --> 00:12:55.320
people there. And they basically said, we want

00:12:55.320 --> 00:12:57.710
to start a non -profit. And we want to call it

00:12:57.710 --> 00:12:58.830
First Draft. So I'm sure we're going to get to

00:12:58.830 --> 00:13:00.889
more of that, but it is really important to remember

00:13:00.889 --> 00:13:03.490
that First Draft really started with a number

00:13:03.490 --> 00:13:05.649
of small number of people who'd been doing this

00:13:05.649 --> 00:13:08.750
work for five or more years. And it was like,

00:13:08.909 --> 00:13:11.250
well, how can we amplify that kind of work? But

00:13:11.250 --> 00:13:13.350
it's really important to remember the foundations

00:13:13.350 --> 00:13:15.629
came from and like Craig Silver was in that group.

00:13:15.750 --> 00:13:17.809
Like it was this amazing group of people that

00:13:17.809 --> 00:13:20.190
were thinking about verification and from different

00:13:20.190 --> 00:13:23.730
perspectives. Yeah. And that was the Google ideas.

00:13:23.750 --> 00:13:25.679
I think they called it at the time. was part

00:13:25.679 --> 00:13:27.399
of that as well. They were very keen to support

00:13:27.399 --> 00:13:29.440
that. I remember a lot of excitement in Google

00:13:29.440 --> 00:13:31.200
about how their tools were being used for the

00:13:31.200 --> 00:13:33.820
human rights thing and then ISIS came along and

00:13:33.820 --> 00:13:35.799
then they got very worried about how their tools

00:13:35.799 --> 00:13:38.159
were being used for other things as well. And

00:13:38.159 --> 00:13:40.299
that's when, especially Facebook, I think as

00:13:40.299 --> 00:13:42.179
well, you've got a lot of concern. Anytime we

00:13:42.179 --> 00:13:43.970
tried to speak to Facebook in that period, it

00:13:43.970 --> 00:13:45.350
was like, oh, we'll have to talk to our lawyers

00:13:45.350 --> 00:13:47.009
about this. It was like, we don't want to engage

00:13:47.009 --> 00:13:49.129
with it. Google are always a lot kind of more,

00:13:49.129 --> 00:13:51.309
I think, communicative. I think they still have

00:13:51.309 --> 00:13:53.309
their, you know, do goods idea, which kind of

00:13:53.309 --> 00:13:55.669
obviously, Tamiya's laser is a bit different,

00:13:56.450 --> 00:13:59.070
but. Yeah, so let's talk about First Draft then

00:13:59.070 --> 00:14:02.629
because I remember meeting people like Ed Weiss

00:14:02.629 --> 00:14:06.409
from Medan and him trying to do work and all

00:14:06.409 --> 00:14:08.289
these different news organizations coming along,

00:14:08.350 --> 00:14:10.669
all these kind of independent and unusual internet

00:14:10.669 --> 00:14:14.750
ones. But we still didn't have that kind of,

00:14:14.750 --> 00:14:18.389
I guess we hadn't really written a lot of stuff

00:14:18.389 --> 00:14:20.710
down that we could show of people. I think that's

00:14:20.710 --> 00:14:22.169
something that First Draft did a really good

00:14:22.169 --> 00:14:25.889
job of kind of trying to start off with. Yeah,

00:14:26.110 --> 00:14:28.070
and at the very beginning, I mean, Google gave

00:14:28.070 --> 00:14:30.049
us an amount of money. It wasn't that much and

00:14:30.049 --> 00:14:32.230
said build a website, which was kind of terrifying

00:14:32.230 --> 00:14:33.769
because nobody, you never want to build a website

00:14:33.769 --> 00:14:36.450
for Google. But what we decided to do as a group,

00:14:36.610 --> 00:14:39.230
as a coalition, is that we said every week we'll

00:14:39.230 --> 00:14:41.730
write one blog post, right? And so you'd write

00:14:41.730 --> 00:14:43.409
something about something you've done, I'd write

00:14:43.409 --> 00:14:44.830
something and then Craig Silverman would write

00:14:44.830 --> 00:14:46.370
something and then me down would write something.

00:14:46.799 --> 00:14:49.580
And I am a little bit biased, but at some point

00:14:49.580 --> 00:14:52.799
somebody will go back to those early first draft

00:14:52.799 --> 00:14:55.659
writings and see it as an archive. When you look

00:14:55.659 --> 00:14:57.159
at everything that was published on first draft,

00:14:57.340 --> 00:15:00.059
it's this kind of incredible six -year journey

00:15:00.059 --> 00:15:02.159
of the ways in which we were all thinking about

00:15:02.159 --> 00:15:05.399
the problem. There was a lot of stuff about definitions,

00:15:05.720 --> 00:15:06.980
and then there was a lot of stuff about bots,

00:15:07.120 --> 00:15:08.360
and then there was a lot of stuff about Russia,

00:15:08.480 --> 00:15:09.940
and then there was a lot of stuff about monetization,

00:15:09.960 --> 00:15:11.360
and then there was a lot of stuff about pre -bunking.

00:15:11.500 --> 00:15:15.200
You can trace the ways in which this field shifted

00:15:15.200 --> 00:15:17.879
based on what people were thinking about at any

00:15:17.879 --> 00:15:20.279
given time. It is kind of extraordinary to go

00:15:20.279 --> 00:15:22.720
back and think about it as an archive of exactly

00:15:22.720 --> 00:15:25.679
the conversation that we're having today. I think

00:15:25.679 --> 00:15:28.500
then 2016 was obviously a big moment because

00:15:28.500 --> 00:15:32.379
of the exit of the Trump election. And Uterte

00:15:32.379 --> 00:15:35.740
in the Philippines. Bellingcat, I was busy with

00:15:35.740 --> 00:15:38.220
all my Syria stuff with Russia starting to bomb

00:15:38.220 --> 00:15:40.440
Syria after we've done the MH17 investigation.

00:15:41.639 --> 00:15:45.100
I remember around that time, we were also getting

00:15:45.100 --> 00:15:47.799
invites to various things. I remember this Facebook

00:15:47.799 --> 00:15:50.039
Paris event where they were trying to convince

00:15:50.039 --> 00:15:51.940
a bunch of fact checkers to work for basically

00:15:51.940 --> 00:15:54.539
free to fact check their website for them. I

00:15:54.539 --> 00:15:56.100
kind of left that meeting thinking I'm not going

00:15:56.100 --> 00:15:58.740
to do that with Burning Cat. But what kind of

00:15:58.740 --> 00:16:00.240
thing were you doing around that period in the

00:16:00.240 --> 00:16:04.259
kind of aftermath of those events? Yeah. So,

00:16:04.659 --> 00:16:07.159
so I came on board full time by that time, first

00:16:07.159 --> 00:16:08.700
draft, I think had gone from this kind of like

00:16:08.700 --> 00:16:10.970
cute hobby. website where we all write one thing

00:16:10.970 --> 00:16:13.450
a week to hang on, there's something more here

00:16:13.450 --> 00:16:15.350
and people need training and we thought about

00:16:15.350 --> 00:16:18.269
training materials. But just after the US election,

00:16:18.549 --> 00:16:20.789
we heard from French journalists who said, well,

00:16:20.809 --> 00:16:22.669
we're worried about what might happen in our

00:16:22.669 --> 00:16:25.730
April election, April 2017. And so a group of

00:16:25.730 --> 00:16:27.669
people got together in Paris, lots of different

00:16:27.669 --> 00:16:30.730
journalists and tech companies and other things.

00:16:30.909 --> 00:16:32.950
And we kind of came up with this idea called

00:16:32.950 --> 00:16:34.830
CrossCheck, which was that these different newsrooms

00:16:34.830 --> 00:16:37.649
would work together. The point being, which is

00:16:37.649 --> 00:16:38.970
the collaboration that we see at Bellingham.

00:16:38.830 --> 00:16:41.769
at which is nobody should be competing. around

00:16:41.769 --> 00:16:44.330
clearing up the shit on the internet. There's

00:16:44.330 --> 00:16:46.250
reasons that journalists compete and it's around

00:16:46.250 --> 00:16:48.269
quality information and it's about investigations,

00:16:48.950 --> 00:16:51.250
but debunking and just cleaning up the mess,

00:16:51.610 --> 00:16:53.250
nobody should be competing about it. So the idea

00:16:53.250 --> 00:16:56.350
was what if journalists actually worked collaboratively

00:16:56.350 --> 00:16:59.190
to write one piece that they could all basically

00:16:59.190 --> 00:17:01.509
syndicate and all publish so they would be using

00:17:01.509 --> 00:17:03.509
less resources to create the same content. And

00:17:03.509 --> 00:17:05.269
then you would also have this amplification effect

00:17:05.269 --> 00:17:07.349
because the same stuff would be appearing. Those

00:17:07.349 --> 00:17:09.690
of us who study Russian disinformation, that's

00:17:09.690 --> 00:17:12.289
what works. firehose of propaganda is that you

00:17:12.289 --> 00:17:15.710
need to be repetitive and actually quality information

00:17:15.710 --> 00:17:17.769
sources often don't do that, like we are competing.

00:17:18.269 --> 00:17:20.369
So that was Cross Check France and that happened

00:17:20.369 --> 00:17:22.950
in 2017. At the end of the election, everyone

00:17:22.950 --> 00:17:25.049
was like, thanks so much, no more funding. We

00:17:25.049 --> 00:17:27.210
did another project like that in Brazil in 2018

00:17:27.210 --> 00:17:29.289
called Comprova, which still continues. I mean,

00:17:29.289 --> 00:17:31.210
I think it's one of the things that I'm most

00:17:31.210 --> 00:17:33.990
proud of is that that coalition of major newsrooms

00:17:33.990 --> 00:17:37.690
in Brazil still works together to counter bad

00:17:37.690 --> 00:17:40.329
information in Brazil. So like many things, Brazil

00:17:40.329 --> 00:17:42.670
have got it right and we've got it wrong. But

00:17:42.670 --> 00:17:45.490
yeah, so first off was ever a fact checking organization,

00:17:45.609 --> 00:17:48.470
but it was how can we connect the dots and train

00:17:48.470 --> 00:17:52.170
people and understand that collaborating around

00:17:52.170 --> 00:17:54.089
these issues is actually a much better model.

00:17:54.089 --> 00:17:56.829
and competing. And what was lovely is that you

00:17:56.829 --> 00:17:58.309
would have these different French journalists

00:17:58.309 --> 00:18:01.079
together. in Slack. I mean, this is another amazing

00:18:01.079 --> 00:18:02.660
archive because the French are so philosophical.

00:18:02.740 --> 00:18:04.319
They'd be like, but what is the meaning of truth?

00:18:04.400 --> 00:18:06.539
Like all in French, like just these great conversations

00:18:06.539 --> 00:18:09.079
about like what they were doing. Um, but you

00:18:09.079 --> 00:18:10.440
also saw journalists from different newsrooms

00:18:10.440 --> 00:18:12.059
were like, oh, you guys do that. Oh, that's interesting.

00:18:12.200 --> 00:18:14.539
We do that. Like, and it made me realize how

00:18:14.539 --> 00:18:17.579
rarely do journalists compare notes. And also

00:18:17.579 --> 00:18:19.839
the quality of the work was higher. This is what

00:18:19.839 --> 00:18:21.740
we found out after was in the evaluation because

00:18:21.740 --> 00:18:23.339
it was said we didn't want to screw up in front

00:18:23.339 --> 00:18:26.259
of our competitors. So the work they were producing

00:18:26.259 --> 00:18:29.259
was of a higher quality, right? So it was a fascinating

00:18:29.259 --> 00:18:31.480
way of thinking about collaboration as a model,

00:18:31.799 --> 00:18:33.460
which I don't think has taken off in the way

00:18:33.460 --> 00:18:35.500
it should have done. But yeah, it was an exciting

00:18:35.500 --> 00:18:38.839
time. I think I often found with those projects,

00:18:39.299 --> 00:18:41.119
again, it's that funding issue. They'll fund

00:18:41.119 --> 00:18:43.380
you for one project, but beyond that, they don't

00:18:43.380 --> 00:18:45.960
offer the support and they just aren't interested

00:18:45.960 --> 00:18:48.400
in sustaining those kinds of projects in the

00:18:48.400 --> 00:18:51.059
long term. And I think looking at what's happening

00:18:51.059 --> 00:18:53.910
in America as well, Thinking about how that collaboration

00:18:53.910 --> 00:18:55.910
is so important, I would like to see a lot more

00:18:55.910 --> 00:18:57.670
of that happening in the US at the moment. And

00:18:57.670 --> 00:19:00.650
it's kind of a shame that America, I don't know

00:19:00.650 --> 00:19:02.309
if they feel that they're kind of special compared

00:19:02.309 --> 00:19:03.970
to the rest of the world, but you know, we've

00:19:03.970 --> 00:19:05.990
seen this happen before in a lots of other contexts.

00:19:06.210 --> 00:19:08.890
We have processes and ideas and we understand

00:19:08.890 --> 00:19:10.750
why building communities around verification

00:19:10.750 --> 00:19:13.069
is important. I feel those lessons need to be

00:19:13.069 --> 00:19:15.569
learned a lot more quickly in this current moment.

00:19:16.960 --> 00:19:18.819
Yeah. Well, I'll just say, going back to the

00:19:18.819 --> 00:19:21.119
funding, is that I'd often say to funders, why

00:19:21.119 --> 00:19:22.680
do you think that disinformation only starts

00:19:22.680 --> 00:19:24.660
12 weeks before the election and then stops a

00:19:24.660 --> 00:19:27.619
day after elections? Why are you funding Brazil,

00:19:27.700 --> 00:19:29.339
then you fund Nigeria, then you fund France,

00:19:29.420 --> 00:19:30.859
then you fund Germany, then you fund the UK,

00:19:31.259 --> 00:19:34.359
as if none of this is connected, right? So I

00:19:34.359 --> 00:19:36.099
was very disappointed by that. And then your

00:19:36.099 --> 00:19:37.960
point, Elliot, is every time there was an election

00:19:37.960 --> 00:19:40.940
in the US, I tried to get funding to do projects

00:19:40.940 --> 00:19:42.480
in the US and never could because it was like,

00:19:42.480 --> 00:19:45.109
oh, we're so different. And of course they're

00:19:45.109 --> 00:19:47.509
not. But there is definitely a superiority complex.

00:19:47.650 --> 00:19:49.049
I say that as somebody who lives here and is

00:19:49.049 --> 00:19:51.369
now a citizen. But yes, it's frustrating that

00:19:51.369 --> 00:19:56.430
there's not more collaboration in the US. I think

00:19:56.430 --> 00:19:59.609
at this point, this would take us to about 2018,

00:19:59.710 --> 00:20:02.230
I think. And is this about the time you joined

00:20:02.230 --> 00:20:07.000
Brown? Is that correct? No, at that point, I

00:20:07.000 --> 00:20:10.119
was still leading First Draft full -time. We

00:20:10.119 --> 00:20:12.039
went to Harvard for a year to try and get funding.

00:20:12.160 --> 00:20:14.400
I mean, it was all this like moving around trying

00:20:14.400 --> 00:20:17.240
to figure out where we could be safe and what

00:20:17.240 --> 00:20:20.460
would work. But yes, it was a real challenge.

00:20:20.660 --> 00:20:24.339
And then really COVID happened and we got significant

00:20:24.339 --> 00:20:26.299
funding from Google, it will surprise nobody.

00:20:26.980 --> 00:20:29.099
A lot of money that was like, you have to spend

00:20:29.099 --> 00:20:31.349
that money in a year. So we kind of expanded

00:20:31.349 --> 00:20:34.750
and did all this work in 2020. And then that

00:20:34.750 --> 00:20:37.109
money went away in 2021 and we had to lay off

00:20:37.109 --> 00:20:41.250
lots of staff. And at that point, you know, attacks

00:20:41.250 --> 00:20:43.150
against those of us that were doing this work

00:20:43.150 --> 00:20:45.049
were increasing. We couldn't get publishing insurance

00:20:45.049 --> 00:20:47.390
because people like we can't like we can't support

00:20:47.390 --> 00:20:49.230
a site that publishes misinformation because

00:20:49.230 --> 00:20:51.500
you're going to get sued. And so it just became

00:20:51.500 --> 00:20:53.420
increasingly difficult for us to do that work.

00:20:53.880 --> 00:20:56.660
And so at that point, I took some staff to Brown

00:20:56.660 --> 00:20:58.640
University, but it was really kind of the end

00:20:58.640 --> 00:21:01.559
of first draft. And I think, Charlie, you put

00:21:01.559 --> 00:21:06.000
the link in. And then lovely Internet Archive,

00:21:06.079 --> 00:21:07.799
who of course we love and we all use the Wayback

00:21:07.799 --> 00:21:10.119
Machine and we have their things. I contacted

00:21:10.119 --> 00:21:12.140
them and said, look, there's such important work

00:21:12.140 --> 00:21:14.400
on here. Is there any way you could help? And

00:21:14.400 --> 00:21:17.799
they were like, oh yeah, we'll archive in perpetuity.

00:21:18.240 --> 00:21:21.259
And so I had this wonderful couple of weeks where

00:21:21.259 --> 00:21:23.180
I had to go through all of the archives of First

00:21:23.180 --> 00:21:26.599
Draft and then pick out how to reorganize the

00:21:26.599 --> 00:21:29.359
site to make it a place that people could go

00:21:29.359 --> 00:21:32.819
and find information. But like I say, it does

00:21:32.819 --> 00:21:34.619
sit to me like an archive that somebody one day

00:21:34.619 --> 00:21:36.140
will go through and study because there's an

00:21:36.140 --> 00:21:38.460
extraordinary depth and wealth of materials,

00:21:38.740 --> 00:21:40.640
both articles and blog posts, but also training

00:21:40.640 --> 00:21:43.359
materials that were translated into many, many

00:21:43.359 --> 00:21:47.319
languages. So yeah, I will always feel very warm

00:21:47.319 --> 00:21:48.920
and fuzzy about First Draft because not only

00:21:48.920 --> 00:21:50.799
was it started by a really great group of people,

00:21:50.880 --> 00:21:52.920
we ended up I think doing work that really helped.

00:21:54.539 --> 00:21:58.539
I think that period around COVID is one where

00:21:58.539 --> 00:22:01.359
now with the rise of this censorship industrial

00:22:01.359 --> 00:22:05.740
complex theory has become a real point of contention.

00:22:05.900 --> 00:22:09.740
I think that was a moment where traditional institutions

00:22:09.740 --> 00:22:13.200
were faced with a huge crisis. They struggled

00:22:13.200 --> 00:22:15.140
to adapt because they were still thinking of

00:22:15.140 --> 00:22:17.519
it in terms of their old institutional dominance

00:22:17.519 --> 00:22:20.400
of the information environment. They were relying

00:22:20.400 --> 00:22:23.259
a lot on academics and independent organizations

00:22:23.259 --> 00:22:26.299
to fill in the gaps that they weren't able to

00:22:26.299 --> 00:22:29.279
handle. I think as well post -2016, there was

00:22:29.279 --> 00:22:31.480
so much focus on the idea of bad information

00:22:31.480 --> 00:22:35.480
versus good information. and that the kind of

00:22:35.480 --> 00:22:37.359
foreign interference had become such a big part

00:22:37.359 --> 00:22:40.700
of the discourse, that really the more underlying

00:22:40.700 --> 00:22:43.640
factors were missing there. I'd be here to hear

00:22:43.640 --> 00:22:45.839
your kind of experiences and reflections in that

00:22:45.839 --> 00:22:48.519
specific period, because that then came back,

00:22:48.599 --> 00:22:50.640
I think, to haunt a lot of us, you know, in the

00:22:50.640 --> 00:22:54.519
new Trump administration. Yeah, no, I have a

00:22:54.519 --> 00:22:56.880
lot to say about this. mean, I'm so glad we're

00:22:56.880 --> 00:22:59.019
having this conversation. But I also have an

00:22:59.019 --> 00:23:01.539
eye to these conversations aren't happening in

00:23:01.539 --> 00:23:04.339
public because there's this sense of if we were

00:23:04.339 --> 00:23:06.759
to kind of acknowledge some of the missteps,

00:23:06.839 --> 00:23:08.400
it would look like that we were kind of giving

00:23:08.400 --> 00:23:11.180
oxygen to certain ideas. But I look back at COVID.

00:23:11.319 --> 00:23:13.680
I mean, even the first draft team, the majority

00:23:13.680 --> 00:23:15.940
of people in their 20s or early 30s, the majority

00:23:15.940 --> 00:23:18.859
of people were living alone. hadn't had a cuddle,

00:23:19.039 --> 00:23:21.660
myself included. I look back now and we were

00:23:21.660 --> 00:23:23.779
all working full time monitoring conspiracy theories

00:23:23.779 --> 00:23:27.019
and everybody was trying to do good. Did that

00:23:27.019 --> 00:23:29.059
mean that we labelled stuff as misinformation

00:23:29.059 --> 00:23:32.420
that was a poor use of that term? 100%. There

00:23:32.420 --> 00:23:34.220
was stuff that we labelled as misinformation

00:23:34.220 --> 00:23:36.579
when we should have said this is unsettled science,

00:23:37.180 --> 00:23:39.920
or we don't have enough evidence yet, or we're

00:23:39.920 --> 00:23:41.339
concerned that people are pushing this narrative

00:23:41.339 --> 00:23:43.779
to make money. you know, we need more scientific.

00:23:43.880 --> 00:23:46.180
We didn't do that, right? People were looking

00:23:46.180 --> 00:23:48.660
for, is this misinformation or not? We were trying

00:23:48.660 --> 00:23:51.119
to ensure that the platforms who had created

00:23:51.119 --> 00:23:54.420
these policies, were they actually doing what

00:23:54.420 --> 00:23:55.579
they said they were going to do in the policies?

00:23:56.000 --> 00:23:58.079
So to your note about the censorship industrial

00:23:58.079 --> 00:24:01.009
complex, like... there's a blog post that kind

00:24:01.009 --> 00:24:03.569
of lists the 50 organizations in the censorship

00:24:03.569 --> 00:24:05.549
industrial complex and like first draft was number

00:24:05.549 --> 00:24:07.450
one, right? And I remember reading that, I mean

00:24:07.450 --> 00:24:10.470
like, but what that was, was just like a number

00:24:10.470 --> 00:24:12.170
of people trying to figure out how to do this

00:24:12.170 --> 00:24:14.630
work. But the platforms, they didn't really go

00:24:14.630 --> 00:24:16.829
after the platforms, right? We weren't calling

00:24:16.829 --> 00:24:19.910
up Google or Twitter, you need to take this down

00:24:19.910 --> 00:24:21.009
because we think it's bad information. We're

00:24:21.009 --> 00:24:23.829
like, hey, your policy says X and here's an example

00:24:23.829 --> 00:24:25.690
of something and we're just letting you know,

00:24:25.710 --> 00:24:27.390
it's up to you to do whatever you want, but this

00:24:27.390 --> 00:24:29.650
goes against your policies. And that's got lost

00:24:29.650 --> 00:24:31.630
in all of this, right? Like it was, it's about

00:24:31.630 --> 00:24:34.309
these individual evil academics or nonprofits

00:24:34.309 --> 00:24:36.890
as opposed to actually these were the platforms

00:24:36.890 --> 00:24:38.829
that were making these kinds of decisions. But

00:24:38.829 --> 00:24:41.089
I do look back and I think, I wish we hadn't

00:24:41.089 --> 00:24:43.309
been moving at speed. I wish we'd had more time

00:24:43.309 --> 00:24:44.589
to collaborate, but we couldn't because we were

00:24:44.589 --> 00:24:47.509
all living alone or not able to connect. And,

00:24:47.809 --> 00:24:49.869
you know, I think that we didn't think through

00:24:49.869 --> 00:24:53.650
that this would look like, we didn't think about

00:24:53.650 --> 00:24:56.410
the balance of freedom of expression. like which

00:24:56.410 --> 00:24:59.329
we all believe in, and how do we ensure freedom

00:24:59.329 --> 00:25:01.609
of expression exists when we have speech that

00:25:01.609 --> 00:25:03.990
is potentially harmful, right? And we still haven't

00:25:03.990 --> 00:25:05.990
figured out that balance, right? Policymakers,

00:25:06.529 --> 00:25:07.910
you know, international human rights lawyers

00:25:07.910 --> 00:25:10.529
haven't figured that out either. And so I think

00:25:10.529 --> 00:25:12.990
organizations like First Draft and other nonprofits

00:25:12.990 --> 00:25:14.529
in this space like Center for Countering Digital

00:25:14.529 --> 00:25:16.250
Hate that are getting absolutely hammered, like

00:25:16.250 --> 00:25:18.309
all of us were trying to do the right thing.

00:25:18.680 --> 00:25:21.039
Was it perfect? Were mistakes, of course, but

00:25:21.039 --> 00:25:22.380
there were mistakes made by all sorts of people

00:25:22.380 --> 00:25:24.200
in all sorts of ways. But we were really good

00:25:24.200 --> 00:25:25.900
people trying to do the right thing, right? But

00:25:25.900 --> 00:25:28.259
we've been painted as these like evil monsters.

00:25:28.660 --> 00:25:33.660
And that's pretty hard to swallow. I think a

00:25:33.660 --> 00:25:35.660
big problem with this as well is the failure

00:25:35.660 --> 00:25:37.920
to recognize the change in the information system

00:25:37.920 --> 00:25:41.759
and how freedom of speech is built around information

00:25:41.759 --> 00:25:43.839
systems that were dominated by institutions,

00:25:44.079 --> 00:25:47.519
by governments, by media, that pre -filtered

00:25:47.519 --> 00:25:50.119
certain views and ideas out of discourse. The

00:25:50.119 --> 00:25:52.160
guy standing on the corner of the street, holding

00:25:52.160 --> 00:25:54.460
a sign saying the earth is flat, didn't get onto

00:25:54.460 --> 00:25:57.799
the news at one BBC. You live in a different

00:25:57.799 --> 00:26:01.559
information environment now. In my work on verification,

00:26:01.700 --> 00:26:04.039
depression and accountability, I like to make

00:26:04.039 --> 00:26:06.980
the point that those processes that were done

00:26:06.980 --> 00:26:09.140
by institutions have kind of been collapsed into

00:26:09.140 --> 00:26:11.180
the public space now. The public aren't just

00:26:11.180 --> 00:26:13.160
recipients of information, we're distributors

00:26:13.160 --> 00:26:16.420
of information, yet we don't hold the responsibility

00:26:16.420 --> 00:26:18.539
of making sure that information is accurate.

00:26:18.680 --> 00:26:20.819
We can just share what the hell we want. I think

00:26:20.819 --> 00:26:23.160
that's where that tension starts being created

00:26:23.160 --> 00:26:25.779
because if you start saying, oh, well, people

00:26:25.779 --> 00:26:28.019
can't say this because they can't say that because

00:26:28.019 --> 00:26:29.940
it's wrong and it's untruthful. Even if you can

00:26:29.940 --> 00:26:32.619
prove that's wrong and untruthful, it's still

00:26:32.619 --> 00:26:36.079
you limiting their speech. But as an academic,

00:26:36.319 --> 00:26:38.680
as someone working at Ballincat, we aren't the

00:26:38.680 --> 00:26:40.799
government, which is really a lot of what freedom

00:26:40.799 --> 00:26:42.940
of speech in the 20th century is focused on.

00:26:43.119 --> 00:26:45.549
Yet we have a problem now that governments are

00:26:45.549 --> 00:26:47.670
trying to adapt to that. So they kind of use

00:26:47.670 --> 00:26:50.309
academics and independent civil society organizations

00:26:50.309 --> 00:26:53.609
as a kind of proxy, and then use that to also

00:26:53.609 --> 00:26:55.609
pressure the social media companies into action.

00:26:55.690 --> 00:26:58.329
And I think there is a legitimate criticism of

00:26:58.329 --> 00:27:00.349
that. I think the censorship industrial complex

00:27:00.349 --> 00:27:03.309
reaches too far. And I think it's very biased

00:27:03.309 --> 00:27:04.829
in the way it's kind of approaching these topics.

00:27:04.829 --> 00:27:07.150
And I think it imagines stronger connections

00:27:07.150 --> 00:27:10.190
between our organizations that really, then it

00:27:10.190 --> 00:27:11.609
actually exists. It's not like we're on the phone

00:27:11.609 --> 00:27:13.529
all the time saying how we censor this because

00:27:13.529 --> 00:27:17.460
we don't like it. Yeah. And so for me, I think

00:27:17.460 --> 00:27:20.339
there is a legitimate criticism of that. I don't

00:27:20.339 --> 00:27:22.059
think the censorship industrial complex theory

00:27:22.059 --> 00:27:24.660
is that legitimate criticism. And it's now being

00:27:24.660 --> 00:27:28.380
weaponized against those people who certain groups

00:27:28.380 --> 00:27:31.579
found, you know, are basically part of the out

00:27:31.579 --> 00:27:33.319
group and the in groups now angry with them.

00:27:33.700 --> 00:27:35.220
I think in this current administration, that's

00:27:35.220 --> 00:27:38.769
become particularly dangerous. Yeah, and I mean,

00:27:38.769 --> 00:27:40.630
I think you get tired if you call out hypocrisy

00:27:40.630 --> 00:27:43.089
all the time. But exactly to your point, you

00:27:43.089 --> 00:27:45.029
know, when we think that's when people say to

00:27:45.029 --> 00:27:46.329
me, oh, yeah, but we just need to change Section

00:27:46.329 --> 00:27:48.349
230. It's like, can we think about what these

00:27:48.349 --> 00:27:50.430
different policies look like in the hands of

00:27:50.430 --> 00:27:52.549
different types of governments? right? Like this

00:27:52.549 --> 00:27:54.410
idea that there are like neutral policies without

00:27:54.410 --> 00:27:56.309
recognizing, well, who does make these decisions

00:27:56.309 --> 00:27:59.329
around speech? And so I think it's very frustrating

00:27:59.329 --> 00:28:01.890
to see, you know, those of us being attacked

00:28:01.890 --> 00:28:03.549
as being part of the censorship industrial, you

00:28:03.549 --> 00:28:05.789
know, I was being investigated by Jim Jordan

00:28:05.789 --> 00:28:08.210
and having my email sent to Congress, like seeing

00:28:08.210 --> 00:28:11.369
all of all of that about being deemed a censor,

00:28:11.470 --> 00:28:14.049
yet seeing like book bans in Florida, like, of

00:28:14.049 --> 00:28:15.470
course, it doesn't make any sense, right? That

00:28:15.470 --> 00:28:17.069
we're not thinking about the system as a whole.

00:28:17.390 --> 00:28:18.910
All of this has been weaponized for political

00:28:18.910 --> 00:28:22.490
gain. And so But I think those of us who got

00:28:22.490 --> 00:28:25.509
into this space were just internet nerds, trying

00:28:25.509 --> 00:28:27.950
to do the right thing. And then I have these

00:28:27.950 --> 00:28:29.769
moments of just like, how on earth did this happen?

00:28:29.869 --> 00:28:32.289
How did I go from working with the BBC about

00:28:32.289 --> 00:28:35.210
how to verify what Beryl from Swansea is sending

00:28:35.210 --> 00:28:38.450
in to being investigated by Jim Jordan? But that's

00:28:38.450 --> 00:28:41.109
this insane journey we've been on. But we have

00:28:41.109 --> 00:28:42.490
to understand it, to your point, Elliot, from

00:28:42.490 --> 00:28:45.920
this very much higher level zoom out of like,

00:28:46.059 --> 00:28:48.019
it's not about the debunking individual claims

00:28:48.019 --> 00:28:49.819
or what was labeled or what wasn't. It's like,

00:28:49.859 --> 00:28:52.019
what's happening at the institutional and ecosystem

00:28:52.019 --> 00:28:54.079
level that really explains all of these changes?

00:28:55.599 --> 00:28:57.880
And I think as we're moving into this current

00:28:57.880 --> 00:28:59.440
period, I mean, as you've mentioned, you've been

00:28:59.440 --> 00:29:02.099
targeted. I think a lot of our colleagues, you

00:29:02.099 --> 00:29:04.200
know, people we've worked with in the past, people

00:29:04.200 --> 00:29:06.200
who are just part of the kind of counter disinformation

00:29:06.200 --> 00:29:08.220
community are finding an incredible amount of

00:29:08.220 --> 00:29:10.799
pressure. I mean, we all know about the MIT technology

00:29:10.799 --> 00:29:13.799
reviewer report where emails have been sent out

00:29:13.799 --> 00:29:16.400
asking for communications from certain individuals.

00:29:16.980 --> 00:29:18.700
You know, Ename and Balincat was also mentioned

00:29:18.700 --> 00:29:22.460
within that. And that in itself is a potentially...

00:29:22.319 --> 00:29:24.420
you know, it pushes people out of the discourse.

00:29:24.579 --> 00:29:28.380
It makes it dangerous. How have you kind of felt,

00:29:28.380 --> 00:29:32.059
you know, in the last year with the kind of rising

00:29:32.059 --> 00:29:36.400
tide of this, your feelings of still being in

00:29:36.400 --> 00:29:38.299
the US, especially when things seem to be getting

00:29:38.299 --> 00:29:41.720
worse rather than getting better? Yeah, I mean,

00:29:41.799 --> 00:29:43.839
I'd be lying if I didn't think that I've had

00:29:43.839 --> 00:29:46.660
like wobbly moments. And I certainly before I

00:29:46.660 --> 00:29:50.019
was a citizen felt much more wobbly. And I mean,

00:29:50.220 --> 00:29:51.759
the reason that these things work is that then

00:29:51.759 --> 00:29:53.779
you have self -censorship, right? Like I was

00:29:53.779 --> 00:29:55.900
less likely to stand on stage and say certain

00:29:55.900 --> 00:29:57.720
things. I was less likely to write blog posts.

00:29:58.200 --> 00:30:00.680
I think now I've got to the point of like, I

00:30:00.680 --> 00:30:02.680
don't want to hide anymore. And, you know, and

00:30:02.680 --> 00:30:05.059
just, but I mean, I think, and it's hard to,

00:30:05.140 --> 00:30:06.799
I mean, I'm sure you get horrific emails. I mean,

00:30:06.920 --> 00:30:09.140
I get horrific emails, which say horrific things

00:30:09.140 --> 00:30:11.019
to me. And there were times I was like, I just

00:30:11.019 --> 00:30:14.700
don't want to deal with it. But I think I've

00:30:14.700 --> 00:30:17.500
come through another side now and just like,

00:30:17.519 --> 00:30:19.640
sod it. I've got a lot to add. And I think it's,

00:30:19.660 --> 00:30:21.660
but I also just want to acknowledge that. Um,

00:30:21.839 --> 00:30:24.200
women in particular get harassed in pretty nasty

00:30:24.200 --> 00:30:26.900
ways. Um, but I do, uh, you know, I know it's

00:30:26.900 --> 00:30:29.339
trite to say, but it's too easy to let them win,

00:30:29.339 --> 00:30:32.140
um, by not putting a head above the parapet and

00:30:32.140 --> 00:30:34.759
saying the things that you need to say. And I

00:30:34.759 --> 00:30:36.500
think it's difficult to have conversations about

00:30:36.500 --> 00:30:38.940
this in public because on one side, you have

00:30:38.940 --> 00:30:40.839
people are really just trying to push a specific

00:30:40.839 --> 00:30:43.059
agenda that, you know, we're the sensors and

00:30:43.059 --> 00:30:45.220
the censorship industrial complexes is evil.

00:30:45.630 --> 00:30:47.150
And if you try and kind of meet them halfway,

00:30:47.250 --> 00:30:48.809
they aren't interested in that because they're

00:30:48.809 --> 00:30:50.710
absolutists in their positions. And I think there

00:30:50.710 --> 00:30:53.730
should be a debate and discussion about where

00:30:53.730 --> 00:30:56.089
things went wrong. But if it just results in

00:30:56.089 --> 00:30:58.410
us being attacked more, then it's very hard to

00:30:58.410 --> 00:31:00.789
have those discussions in public. And Charlotte,

00:31:00.869 --> 00:31:03.009
how are we doing for questions from the audience?

00:31:04.490 --> 00:31:09.430
A couple of people have asked about AI, actually.

00:31:09.750 --> 00:31:11.809
Rope Stone wrote, if governments are weaponizing

00:31:11.809 --> 00:31:14.700
disinformation to the public, How can the consumer

00:31:14.700 --> 00:31:16.859
in the street stand a chance of finding the truth?

00:31:17.500 --> 00:31:20.599
Will the trend of LLM and AI give way to much

00:31:20.599 --> 00:31:22.539
information for the user to comprehend what is

00:31:22.539 --> 00:31:25.440
actual truth? And as a secondary to that, Sarah

00:31:25.440 --> 00:31:27.960
said, is there any old school verification techniques

00:31:27.960 --> 00:31:33.019
that can sidestep this new trend in people trying

00:31:33.019 --> 00:31:36.519
to use AI to verify things? Yeah, I just taught

00:31:36.519 --> 00:31:39.180
a class just now actually on generative AI, good,

00:31:39.200 --> 00:31:41.559
bad and ugly. And I was trying to say to the

00:31:41.559 --> 00:31:43.440
students like the thing that's the ugliest part

00:31:43.440 --> 00:31:45.619
for me is the fact that people are losing trust

00:31:45.619 --> 00:31:48.000
in anything they see or hear. Right. And we saw

00:31:48.000 --> 00:31:50.039
that the first time I saw that was around Gaza

00:31:50.039 --> 00:31:52.559
when I saw people like that's so upsetting that

00:31:52.559 --> 00:31:55.740
can't possibly be true. But seeing yesterday

00:31:55.740 --> 00:31:59.480
with the Alex Pretty video coming out of like

00:31:59.470 --> 00:32:01.869
you know, him potentially hitting a car and spitting,

00:32:01.930 --> 00:32:03.690
and seeing people like, that can't be true, that

00:32:03.690 --> 00:32:05.250
can't be true. And I was like, yeah, that's the

00:32:05.250 --> 00:32:08.490
biggest victim of this moment. Um, but I think

00:32:08.490 --> 00:32:10.650
what's so interesting about AI verification is,

00:32:10.690 --> 00:32:12.190
of course, there's all these startups trying

00:32:12.190 --> 00:32:13.950
to convince us all that they've got the technology

00:32:13.950 --> 00:32:16.589
that can do this automatically. But in the same

00:32:16.589 --> 00:32:18.789
way as we started when we were doing this work,

00:32:18.789 --> 00:32:21.470
you know, 15 years ago, there was never one easy

00:32:21.470 --> 00:32:23.470
tool, right? You had to, like, look at the exif

00:32:23.470 --> 00:32:25.329
data and then do the shadow analysis and then

00:32:25.329 --> 00:32:27.089
check the timestamp on the, like, there were

00:32:27.089 --> 00:32:28.730
all these different checks you had to make. And

00:32:28.730 --> 00:32:32.009
that's what I say with AI, which is like, the

00:32:32.009 --> 00:32:33.930
stuff that you see, you've still got to do the

00:32:33.930 --> 00:32:35.990
human checks of like, did this, do you have evidence

00:32:35.990 --> 00:32:38.150
that this happened independently? And if you

00:32:38.150 --> 00:32:40.630
don't keep quiet, like there are times that newsrooms

00:32:40.630 --> 00:32:42.609
were like, we've got 99 % there and the newsrooms

00:32:42.609 --> 00:32:44.450
are like, but we're not 100 % there. So we're

00:32:44.450 --> 00:32:46.630
not going to use that footage. But I think we're

00:32:46.630 --> 00:32:49.630
so far away from that now that that's my concern

00:32:49.630 --> 00:32:52.809
about AI. But I do hope that with this new generation

00:32:52.809 --> 00:32:55.710
of students and people who have grown up with

00:32:55.710 --> 00:32:58.369
AI, we also need to keep saying if we all lose

00:32:58.369 --> 00:32:59.970
trust in everything, we're absolutely screwed.

00:33:00.319 --> 00:33:03.200
And B, here's some fun techniques. So when people

00:33:03.200 --> 00:33:05.420
say, well, what's the magic AI verification tool?

00:33:05.500 --> 00:33:09.440
I'm like, here are the magic non -AI verification

00:33:09.440 --> 00:33:11.680
tools that still stand the test of time, which

00:33:11.680 --> 00:33:13.019
is like picking up the phone and talking to somebody

00:33:13.019 --> 00:33:15.480
and be like, were you there? Were you an eyewitness?

00:33:15.900 --> 00:33:18.319
Can you tell me what you saw? Sometimes there's

00:33:18.319 --> 00:33:20.599
just no shortcut. So yeah, I'm interested to

00:33:20.599 --> 00:33:23.900
hear what Elliot says. And I think on that example

00:33:23.900 --> 00:33:26.839
you gave of the recent video of Alex Petty kicking

00:33:26.839 --> 00:33:30.339
the... A lot of people I saw reacting, oh, that's

00:33:30.339 --> 00:33:32.680
AI, that's not real, it's been manipulated. That's

00:33:32.680 --> 00:33:34.539
no different when Trump was showing the pictures

00:33:34.539 --> 00:33:36.900
of the video of the things being thrown out the

00:33:36.900 --> 00:33:38.380
White House window and he said, oh, that's probably

00:33:38.380 --> 00:33:41.160
AI, it's not real, it's AI. It's the same kind

00:33:41.160 --> 00:33:43.140
of mindset that it's about protecting the group

00:33:43.140 --> 00:33:46.339
identity rather than verifying something. And

00:33:46.339 --> 00:33:48.960
if he kicked that car, it doesn't justify him

00:33:48.960 --> 00:33:52.019
being shot to death two weeks later. Not all

00:33:52.019 --> 00:33:54.700
the people using it as justification are either

00:33:54.700 --> 00:33:57.579
bootlegs or psychopaths, I think. So But it is

00:33:57.579 --> 00:34:00.380
about, you know, group identity for a lot of

00:34:00.380 --> 00:34:02.339
people on both sides of these situations. And

00:34:02.339 --> 00:34:04.859
I think we need to commit ourselves primarily

00:34:04.859 --> 00:34:07.559
to, you know, building our group identity around

00:34:07.559 --> 00:34:10.059
that verification, deliberation, accountability

00:34:10.059 --> 00:34:12.300
process being functional rather than serving

00:34:12.300 --> 00:34:15.719
the interests of the group. Yeah. And I do think

00:34:15.719 --> 00:34:17.619
whenever, you know, I talk to students and others,

00:34:17.739 --> 00:34:19.639
there's like, here's the technical verification

00:34:19.639 --> 00:34:22.469
stuff, but here's why our brains are mush. like

00:34:22.469 --> 00:34:25.289
left or right, like this is why we're vulnerable

00:34:25.289 --> 00:34:27.809
to this. And the in -group, out -group, the fact

00:34:27.809 --> 00:34:29.349
that we're all performing when we're on social

00:34:29.349 --> 00:34:31.030
media, that's the stuff that I think doesn't

00:34:31.030 --> 00:34:32.809
get taught as much when we do a checklist of

00:34:32.809 --> 00:34:34.670
like, what's the pope wearing a puffer jacket?

00:34:34.929 --> 00:34:36.369
But yeah, all right, we can do those checks.

00:34:36.409 --> 00:34:39.150
But actually, why might you be convinced? Well,

00:34:39.150 --> 00:34:40.329
because you looked great in that puffer jacket.

00:34:40.329 --> 00:34:41.590
And you want it to be true because it's funny,

00:34:41.630 --> 00:34:43.869
right? Like that's a silly example. But we don't

00:34:43.869 --> 00:34:45.889
talk enough about the psychology of all of this.

00:34:50.790 --> 00:34:53.010
platforms when you were talking and a couple

00:34:53.010 --> 00:34:55.590
of questions have come in that. When did the

00:34:55.590 --> 00:34:58.150
tensions between journalists, researchers and

00:34:58.150 --> 00:35:01.750
platforms really emerge when it comes to disinformation?

00:35:03.150 --> 00:35:05.989
And Nick asked, if governments, can governments

00:35:05.989 --> 00:35:08.630
regulate platforms to make algorithmic curation

00:35:08.630 --> 00:35:14.239
less alien in the attention economy? So the first

00:35:14.239 --> 00:35:16.820
question, I mean, I have to say as somebody who's

00:35:16.820 --> 00:35:20.000
probably been to like 400, 500 counter disinformation

00:35:20.000 --> 00:35:21.639
conferences, I mean, that's where the industrial

00:35:21.639 --> 00:35:23.820
complex was just like event after event after

00:35:23.820 --> 00:35:26.780
event. And people would always invite the platforms

00:35:26.780 --> 00:35:28.260
and were like, yeah, we got Google to come, or

00:35:28.260 --> 00:35:30.219
we got Facebook to come. And the representatives

00:35:30.219 --> 00:35:32.840
would turn up with their marketing slides that

00:35:32.840 --> 00:35:35.539
had been signed off by the... And they said nothing,

00:35:35.599 --> 00:35:36.900
right? But we were like, oh, we should keep inviting

00:35:36.900 --> 00:35:40.360
them. I think the shift happened in 2021 when

00:35:40.360 --> 00:35:43.059
everybody realized post -COVID that there was

00:35:43.059 --> 00:35:46.079
a shift happening and there was fears around...

00:35:46.079 --> 00:35:48.599
There was starting to be these grumblings that

00:35:48.599 --> 00:35:50.579
there had been jaw boning, the idea that the

00:35:50.579 --> 00:35:52.369
Biden administration had been... putting pressure

00:35:52.369 --> 00:35:55.230
on the platforms. And it was around that time

00:35:55.230 --> 00:35:58.130
that people really stopped talking to platforms

00:35:58.130 --> 00:36:00.090
because there was fears of like collusion and

00:36:00.090 --> 00:36:02.730
jawboning. But that has really, I mean, I used

00:36:02.730 --> 00:36:04.369
to speak to almost everybody, like lots of people

00:36:04.369 --> 00:36:07.730
at all platforms a lot of the time and like all

00:36:07.730 --> 00:36:09.730
of those relationships have disappeared. And

00:36:09.730 --> 00:36:11.570
then of course, Musk took over, trust and safety

00:36:11.570 --> 00:36:14.190
teams were demolished. But yeah, again, it's

00:36:14.190 --> 00:36:16.630
kind of very interesting to track like Google

00:36:16.630 --> 00:36:19.539
funding a first draft. And the other day, on

00:36:19.539 --> 00:36:22.420
my reading list, I had a piece of research from

00:36:22.420 --> 00:36:25.300
Jigsaw about toxicity online. And my student

00:36:25.300 --> 00:36:26.739
was like, hey, Claire, I can't find the PDF.

00:36:26.800 --> 00:36:27.780
And I was like, oh, here's a link. And I was

00:36:27.780 --> 00:36:29.780
like, oh, it's been taken down. I mean, Google

00:36:29.780 --> 00:36:32.019
is so frightened that some of their best research

00:36:32.019 --> 00:36:33.360
has been taken down off the internet because

00:36:33.360 --> 00:36:35.320
they don't want to upset Trump, right? So it's

00:36:35.320 --> 00:36:37.880
an extraordinary journey that we've been on.

00:36:38.179 --> 00:36:40.059
The question about regulation is a really, really

00:36:40.059 --> 00:36:43.179
good one. I used to believe that we would get

00:36:43.179 --> 00:36:44.960
someone with regulation. And I was very excited

00:36:44.960 --> 00:36:47.690
about the Digital Services Act from the EU. that

00:36:47.690 --> 00:36:49.829
there was a real emphasis on data transparency,

00:36:50.150 --> 00:36:53.329
a real emphasis on auditing algorithms as opposed

00:36:53.329 --> 00:36:56.670
to content. high hopes. And then of course, as

00:36:56.670 --> 00:36:58.989
we've seen, since Trump has come back into office,

00:36:59.389 --> 00:37:03.969
pressure from the US against the EU on these

00:37:03.969 --> 00:37:06.250
fines. And I worry that there will be some element

00:37:06.250 --> 00:37:08.730
of pulling back and a chilling effect when I

00:37:08.730 --> 00:37:10.949
think if Trump hadn't got in, we would have seen

00:37:10.949 --> 00:37:14.130
like some pretty brave moves from the EU against

00:37:14.130 --> 00:37:16.889
the platforms. I do not see that happening. I

00:37:16.889 --> 00:37:20.050
wish I could be wrong, but I'm not as hopeful

00:37:20.050 --> 00:37:22.070
as I was two years ago about what regulation

00:37:22.070 --> 00:37:26.719
might be able to do. And I would add to that

00:37:26.719 --> 00:37:29.440
that we've had the same experience. I mean, up

00:37:29.440 --> 00:37:31.679
until Musk took over Twitter, I mean, like literally

00:37:31.679 --> 00:37:34.360
as he took over, we were just starting a new

00:37:34.360 --> 00:37:38.159
project with Twitter about looking kind of identifying

00:37:38.159 --> 00:37:41.000
disinformation that works and stuff like that.

00:37:41.380 --> 00:37:43.820
And literally the team's emails stopped working

00:37:43.820 --> 00:37:46.360
overnight once Musk took over. And I had a lot

00:37:46.360 --> 00:37:48.219
of people I knew very well at Twitter who had

00:37:48.219 --> 00:37:49.960
been really working on these issues seriously.

00:37:50.380 --> 00:37:53.000
just kind of disappearing. Like, you know, as

00:37:53.000 --> 00:37:54.739
we all contact each other by email and then their

00:37:54.739 --> 00:37:56.440
emails work, it's like, I don't have their phone

00:37:56.440 --> 00:37:58.260
number or any of the details for them. So yeah,

00:37:58.360 --> 00:38:00.380
and it was, it felt quite sudden when it happened,

00:38:00.599 --> 00:38:02.400
not just in terms of X, but just that, like you

00:38:02.400 --> 00:38:06.789
said, the post 2021 period. We were talking a

00:38:06.789 --> 00:38:09.750
second ago about the U .S. Sarah said, are you

00:38:09.750 --> 00:38:12.449
seeing any changes in U .S. organizations going

00:38:12.449 --> 00:38:14.670
from, no, we don't need any help researching

00:38:14.670 --> 00:38:17.670
disinformation, to we really need help researching

00:38:17.670 --> 00:38:20.030
disinformation? Or are you seeing the opposite

00:38:20.030 --> 00:38:25.110
because of Trump in power? Yeah, I mean, the

00:38:25.110 --> 00:38:27.909
sad thing is, is that kind of three years ago,

00:38:27.949 --> 00:38:30.469
there was a really thriving network of research

00:38:30.469 --> 00:38:32.969
institutes and nonprofits in the US that were

00:38:32.969 --> 00:38:35.630
being funded and supported. And that network

00:38:35.630 --> 00:38:40.170
has almost entirely gone right from the project

00:38:40.170 --> 00:38:43.670
at Stanford with Ray DeResta and Alex Deimos.

00:38:45.489 --> 00:38:48.030
Stanford basically once they were attacked. shutting

00:38:48.030 --> 00:38:50.530
it down, you know, Joan Donovan at Harvard, we

00:38:50.530 --> 00:38:52.550
were doing work at Brown, you know, like it's

00:38:52.550 --> 00:38:54.730
first draft has gone, like it, there's many,

00:38:54.769 --> 00:38:56.789
many, many organizations that could just couldn't

00:38:56.789 --> 00:39:00.150
get any funding anymore, or were basically attacked

00:39:00.150 --> 00:39:02.829
into, and they're bigger organizations saying

00:39:02.829 --> 00:39:04.449
when we, you know, we can't support you anymore.

00:39:04.969 --> 00:39:07.690
So it is extraordinary to me to see the collapse

00:39:07.690 --> 00:39:11.329
of a pretty tight knit. infrastructure of counter

00:39:11.329 --> 00:39:14.230
disinformation efforts in the US completely disappear

00:39:14.230 --> 00:39:16.349
at a time when we need them more. But that was

00:39:16.349 --> 00:39:18.150
the whole purpose. That was the whole purpose

00:39:18.150 --> 00:39:20.389
and it was successful and it stopped people from

00:39:20.389 --> 00:39:21.849
doing the monitoring. You know, we're going to

00:39:21.849 --> 00:39:25.269
go into the midterms. In fact, in the 2024 elections,

00:39:25.809 --> 00:39:27.530
we used to have a whole network of people that

00:39:27.530 --> 00:39:29.170
would do these big election projects and would

00:39:29.170 --> 00:39:31.550
be monitoring. And all of that has disappeared

00:39:31.550 --> 00:39:33.929
now. So we just don't have the monitoring infrastructure

00:39:33.929 --> 00:39:36.230
that we used to have. So yeah, it's pretty shocking.

00:39:37.219 --> 00:39:41.260
I'd add that as well. The cut to USAID funding

00:39:41.260 --> 00:39:43.639
has had a devastating effect outside of the US

00:39:43.639 --> 00:39:48.159
as well. We know loads of media organizations

00:39:48.159 --> 00:39:50.219
who are using USAID funding, a lot were very

00:39:50.219 --> 00:39:52.320
reliant on it. They've all started to collapse

00:39:52.320 --> 00:39:54.920
and disappear. But also that has an impact on

00:39:54.920 --> 00:39:57.780
the wider funding ecosystems. It means there's

00:39:57.780 --> 00:40:00.239
less money in general for organizations who are

00:40:00.239 --> 00:40:02.760
doing this kind of work and priorities are changing

00:40:02.760 --> 00:40:07.269
a bit. And I think also kind of funders are starting

00:40:07.269 --> 00:40:09.489
to feel that maybe this counter disinformation

00:40:09.489 --> 00:40:11.849
stuff, these fact -checking websites aren't actually

00:40:11.849 --> 00:40:14.030
as impactful as they think they are. And I think

00:40:14.030 --> 00:40:17.489
they are correct to an extent on that. There's

00:40:17.489 --> 00:40:19.849
no money now to figure out the more effective

00:40:19.849 --> 00:40:21.989
version of kind of responding to these issues.

00:40:22.889 --> 00:40:24.650
And, you know, that's affecting every organization.

00:40:24.889 --> 00:40:26.309
It certainly affects Bellingcat, even though

00:40:26.309 --> 00:40:28.670
we never took USAID funding, we're still competing

00:40:28.670 --> 00:40:32.099
in that much smaller pool of money. And there's

00:40:32.099 --> 00:40:34.699
also been a huge number of journalists, researchers

00:40:34.699 --> 00:40:37.840
made redundant from these roles. So if you're

00:40:37.840 --> 00:40:40.079
working in it, it's very difficult to find any

00:40:40.079 --> 00:40:42.579
kind of employment. And you are seeing, I think,

00:40:42.639 --> 00:40:45.059
a bit of a rise in more independent outlets,

00:40:45.059 --> 00:40:46.599
but still they've got the challenge of having

00:40:46.599 --> 00:40:49.000
a subscriber base or, you know, hoping Substack

00:40:49.000 --> 00:40:51.780
works with them or competing in this dwindling

00:40:51.780 --> 00:40:58.179
pool of money. I was just going to talk about

00:40:58.179 --> 00:41:00.579
platform money, right? First off, we definitely

00:41:00.579 --> 00:41:02.599
got criticized for taking platform money. My

00:41:02.599 --> 00:41:04.840
perspective was always, we got money to do good

00:41:04.840 --> 00:41:07.480
work, right? And there was always editorial independence.

00:41:07.800 --> 00:41:10.280
And why shouldn't they pay to help clean up the

00:41:10.280 --> 00:41:12.940
mess that they'd created? But even seeing how

00:41:12.940 --> 00:41:15.280
Meta has pulled back, I mean, collapsing fact

00:41:15.280 --> 00:41:17.800
checking in the US. And we can have a longer

00:41:17.800 --> 00:41:19.300
conversation about the efficacy of fact checking.

00:41:19.579 --> 00:41:21.420
But fact checking when you're doing it on Meta's

00:41:21.420 --> 00:41:23.400
platforms and the reach that those fact checks

00:41:23.400 --> 00:41:26.800
got once they retrained the algorithms to basically

00:41:26.760 --> 00:41:28.619
to lower certain types of content or to flag

00:41:28.619 --> 00:41:31.760
others. But yeah, there are many, many fact checkers

00:41:31.760 --> 00:41:34.400
that started because of money from meta that

00:41:34.400 --> 00:41:36.639
have now, in many, many parts of the world, that

00:41:36.639 --> 00:41:41.820
we no longer have fact checkers. Have we in the

00:41:41.820 --> 00:41:44.940
past overestimated the power of kind of corrections

00:41:44.940 --> 00:41:48.019
and fact -checking? Kouch von Katschi asked a

00:41:48.019 --> 00:41:50.019
question that kind of fits in here. How do we

00:41:50.019 --> 00:41:52.519
go about teaching how to navigate this new information

00:41:52.519 --> 00:41:54.980
landscape that is very fractured and has decreasing

00:41:54.980 --> 00:41:57.219
trust? It feels like we're in a massive information

00:41:57.219 --> 00:41:59.739
and geopolitical paradigm shift, and I'm curious

00:41:59.739 --> 00:42:03.019
how we work through this. So if immediate fact

00:42:03.019 --> 00:42:06.059
-checking isn't the way forward now, what would

00:42:06.059 --> 00:42:11.210
you suggest is? So I'll be transparent in that

00:42:11.210 --> 00:42:13.750
I'm a trustee of Full Fact, which is a UK fact

00:42:13.750 --> 00:42:16.829
-checking organisation. I've always been clear

00:42:16.829 --> 00:42:20.110
that fact -checking in terms of changing an individual's

00:42:20.110 --> 00:42:22.489
belief system is very, very difficult. Fact -checking

00:42:22.489 --> 00:42:24.349
is incredibly important and even more so now

00:42:24.349 --> 00:42:26.489
that we have an accurate record. In an age of

00:42:26.489 --> 00:42:28.849
LLMs, it's even more important that we actually

00:42:28.849 --> 00:42:31.949
do have an archival moment that says what something

00:42:31.949 --> 00:42:36.309
was, right? So for me, it's about having an accurate

00:42:36.309 --> 00:42:39.110
record and it's also at times ensuring that the

00:42:39.110 --> 00:42:41.030
person who lied in the first place changes it.

00:42:41.130 --> 00:42:42.429
They're very good at getting people to correct

00:42:42.429 --> 00:42:46.690
the record. it is all really important. But to

00:42:46.690 --> 00:42:49.130
Elliot's point, we spent so much money on the

00:42:49.130 --> 00:42:51.670
end of the information supply chain, right? Like,

00:42:51.670 --> 00:42:53.210
let's wait until people are seeing the bad stuff

00:42:53.210 --> 00:42:54.949
and then we'll kind of issue a correction. We

00:42:54.949 --> 00:42:56.510
should have been working much, much earlier in

00:42:56.510 --> 00:42:59.289
the supply chain. And we didn't do that. But

00:42:59.289 --> 00:43:00.449
that doesn't mean that fact checking doesn't

00:43:00.449 --> 00:43:02.110
work. It just doesn't work in the way that I

00:43:02.110 --> 00:43:03.670
think a lot of policymakers thought was like,

00:43:03.670 --> 00:43:05.329
oh, we just need people to have more facts and

00:43:05.329 --> 00:43:08.230
then they'll change their minds. But it is astonishing

00:43:08.230 --> 00:43:10.670
to me when you look at community notes on X and

00:43:10.670 --> 00:43:12.750
other places, how many times people linked to

00:43:12.750 --> 00:43:15.380
fact checkers. and like the LLMs, like if you

00:43:15.380 --> 00:43:17.980
go to ChatGBT, it pulls all of that stuff from

00:43:17.980 --> 00:43:19.699
those small fact checkers, has never paid a penny

00:43:19.699 --> 00:43:22.079
for that information, uses all of their archives

00:43:22.079 --> 00:43:24.960
to train their models. So for that now, I actually

00:43:24.960 --> 00:43:26.260
think the fact checking is more important than

00:43:26.260 --> 00:43:30.119
ever, but the funding is disappearing. I always

00:43:30.119 --> 00:43:33.340
find it amusing that Grok is a big fan of BedlingCat

00:43:33.340 --> 00:43:36.280
despite Elon Musk's position on the BedlingCat.

00:43:37.820 --> 00:43:40.860
I mean, for me, a lot of this is actually getting

00:43:41.400 --> 00:43:43.679
into that discourse as it's unfolding. And I

00:43:43.679 --> 00:43:45.239
think the recent shootings has been a really

00:43:45.239 --> 00:43:47.159
good example of that. We knew the US government

00:43:47.159 --> 00:43:48.880
was going to lie about the second shooting because

00:43:48.880 --> 00:43:51.500
they lied about the first one. So at Bellingham,

00:43:51.559 --> 00:43:53.619
we were very keen to make sure we could get information

00:43:53.619 --> 00:43:55.800
out very quickly. So we published stuff up through

00:43:55.800 --> 00:43:58.280
Blue Sky and on our other platforms without in,

00:43:58.280 --> 00:44:00.199
you know, hours of the shooting happening as

00:44:00.199 --> 00:44:02.599
that discourse was forming and we still were

00:44:02.599 --> 00:44:04.460
able to do our kind of longer form reporting

00:44:04.460 --> 00:44:06.900
later on. Also training other people to do that's

00:44:06.900 --> 00:44:10.380
really important. You had lots of media, traditional

00:44:10.380 --> 00:44:13.139
media doing open source investigation into that

00:44:13.139 --> 00:44:15.619
incident, partly because they've been trained

00:44:15.619 --> 00:44:17.599
to do that. Banning Cat has trained a lot of

00:44:17.599 --> 00:44:20.360
those people. We've created content online. You've

00:44:20.360 --> 00:44:22.320
done that kind of work. It's put those skills

00:44:22.320 --> 00:44:25.280
out there, but also involving the public. We

00:44:25.280 --> 00:44:28.760
asked a lot about this and say that this Discord

00:44:28.760 --> 00:44:31.659
community that we're talking to now played a

00:44:31.659 --> 00:44:33.559
role in getting that information out. Our trained

00:44:33.559 --> 00:44:36.039
volunteers community had a role in getting that

00:44:36.039 --> 00:44:38.989
information out. And by spreading those skills,

00:44:39.110 --> 00:44:41.050
it means there's more people applying them to

00:44:41.050 --> 00:44:43.550
this information, often in places that you aren't

00:44:43.550 --> 00:44:46.590
looking at. I always think of the role the kind

00:44:46.590 --> 00:44:49.409
of alternative health community played in spreading

00:44:49.409 --> 00:44:52.570
COVID disinformation. They weren't formed around

00:44:52.570 --> 00:44:54.730
COVID disinformation. They were formed about

00:44:54.730 --> 00:44:56.809
what were often traumatic experiences they had

00:44:56.809 --> 00:44:58.989
with medical professionals. And then they built

00:44:58.989 --> 00:45:04.289
a network around that, which towards more disordered

00:45:04.289 --> 00:45:10.530
forms of communication. We can't predict where

00:45:10.530 --> 00:45:12.869
this stuff can come from, so for me it's about

00:45:12.869 --> 00:45:15.429
spreading those skills as far and wide as possible

00:45:15.429 --> 00:45:18.210
through every layer of society so that when this

00:45:18.210 --> 00:45:19.909
stuff comes out, and as I mentioned before, we're

00:45:19.909 --> 00:45:22.170
all distributors of information, people at least

00:45:22.170 --> 00:45:24.010
think twice about the stuff they're sharing,

00:45:24.369 --> 00:45:26.269
and maybe that will stop some of the formation

00:45:26.269 --> 00:45:28.090
of these very strong opinions that are based

00:45:28.090 --> 00:45:32.760
on fictions. Yeah, and I'll just support that.

00:45:32.840 --> 00:45:34.639
Like I make jokes every time I speak about how

00:45:34.639 --> 00:45:36.460
I spend a lot of my days in the armpit of the

00:45:36.460 --> 00:45:38.800
internet. And so I spend a lot of time in spaces

00:45:38.800 --> 00:45:41.280
that are deeply participatory, that they're conspiracy

00:45:41.280 --> 00:45:43.320
communities or anti -vax communities. But they

00:45:43.320 --> 00:45:45.519
are communities where people turn up, they feel

00:45:45.519 --> 00:45:47.340
heard, they're told that they can do their own

00:45:47.340 --> 00:45:49.599
research, they have a sense of agency. Right.

00:45:49.599 --> 00:45:52.599
And the only community to me that matches those

00:45:52.599 --> 00:45:54.480
kind of communities is the Bellingcat community.

00:45:54.739 --> 00:45:57.659
I think if you think about trusted gatekeeping

00:45:57.659 --> 00:46:00.230
type organizations, they're still top down. put

00:46:00.230 --> 00:46:01.889
the right person behind the podium or you put

00:46:01.889 --> 00:46:04.289
out the right PDF, right? There's this idea that

00:46:04.289 --> 00:46:07.289
it's still top -down and hierarchical. And very,

00:46:07.309 --> 00:46:10.449
very few organizations, if any, on the kind of

00:46:10.449 --> 00:46:13.090
the side of quality information have said in

00:46:13.090 --> 00:46:15.710
order to get better information, you need collaborative

00:46:15.710 --> 00:46:18.590
spaces where people feel a sense of agency and

00:46:18.590 --> 00:46:21.050
you can work together, right? So just to say,

00:46:21.510 --> 00:46:23.230
I hope you will realize how important this space

00:46:23.230 --> 00:46:25.510
is, but it really is a model that I think many,

00:46:25.570 --> 00:46:26.869
many organizations, because they feel like they

00:46:26.869 --> 00:46:29.559
lose control if they hand it over. right? If

00:46:29.559 --> 00:46:32.199
they hand over messaging or communication, there's

00:46:32.199 --> 00:46:33.960
this idea that the quality goes down as opposed

00:46:33.960 --> 00:46:35.619
to, well, actually, if you hand it over, the

00:46:35.619 --> 00:46:37.280
quality will go up, but you're going to lose

00:46:37.280 --> 00:46:40.559
control. So I couldn't say more about how important

00:46:40.559 --> 00:46:41.840
this kind of community is. And I saw something

00:46:41.840 --> 00:46:43.400
in the chat being like, yeah, but these kind

00:46:43.400 --> 00:46:46.260
of communities are slower and other things. I

00:46:46.260 --> 00:46:48.920
do ultimately think we're in this really nasty

00:46:48.920 --> 00:46:51.460
moment right now where, like, how many times

00:46:51.460 --> 00:46:53.039
do you hear people go, oh yeah, well, I'm on

00:46:53.039 --> 00:46:55.380
Instagram. I mean, I know it's 75 % AI slop,

00:46:55.400 --> 00:46:57.820
but I still look at it, right? I think in a few

00:46:57.820 --> 00:47:00.000
years time, we'll actually have a kind of a bifurcated

00:47:00.000 --> 00:47:02.460
internet where it's not like I don't love AI

00:47:02.460 --> 00:47:04.500
slop sometimes. Like I want to see a cat singing

00:47:04.500 --> 00:47:06.559
and dancing, right? Because I'm in a bad mood

00:47:06.559 --> 00:47:08.139
and I want to be cheered up. But there are other

00:47:08.139 --> 00:47:09.920
times where I only want good information. And

00:47:09.920 --> 00:47:12.219
I think there will be increasing spaces where

00:47:12.670 --> 00:47:14.610
quality information coming from like quality

00:47:14.610 --> 00:47:18.269
organizations will have, you know, like watermarked

00:47:18.269 --> 00:47:21.349
content or they will have, you know, the kind

00:47:21.349 --> 00:47:23.570
of things that we're seeing around AI generated

00:47:23.570 --> 00:47:25.409
content that says this really was filmed by the

00:47:25.409 --> 00:47:27.550
New York Times or this really was filmed by or

00:47:27.550 --> 00:47:29.909
created by Bellingcat so that you can go to places

00:47:29.909 --> 00:47:32.289
for that kind of content. Doesn't mean we won't

00:47:32.289 --> 00:47:34.389
have the nasty sloppy parts of the internet,

00:47:34.409 --> 00:47:36.369
but I think people will realize that you can't

00:47:36.369 --> 00:47:38.789
have both in the same space. It shouldn't be

00:47:38.789 --> 00:47:40.610
on all of us individually to figure out is that

00:47:40.610 --> 00:47:42.829
true or not. So it could be wrong, but that's

00:47:42.829 --> 00:47:44.369
a kind of a prediction I have for the next five

00:47:44.369 --> 00:47:47.909
years. I'll just add to that. I had it interesting

00:47:47.909 --> 00:47:51.349
when I left X and then into Blue Sky. I wasn't

00:47:51.349 --> 00:47:53.070
planning to join Blue Sky, but to be honest,

00:47:53.090 --> 00:47:56.250
when I quit X, I felt like physically depressed.

00:47:56.510 --> 00:47:57.969
It was like I was so addicted. It's like going

00:47:57.969 --> 00:48:00.110
clawed into me so much. I thought I'd need like

00:48:00.110 --> 00:48:02.010
a methadone to get over this. And that's what

00:48:02.010 --> 00:48:04.150
Blue Sky became for me. But then when I actually

00:48:04.150 --> 00:48:07.469
started posting there, the discourse felt so

00:48:07.469 --> 00:48:10.570
different. There's so few times when someone

00:48:10.570 --> 00:48:12.710
would post something like banning Kat as the

00:48:12.710 --> 00:48:15.630
CIA or how terrible I was, which I got way too

00:48:15.630 --> 00:48:20.230
used to on X. And I think, you know, that having

00:48:20.230 --> 00:48:22.050
those good spaces where there can be discourse,

00:48:22.130 --> 00:48:23.789
where there is engagement, that you can talk

00:48:23.789 --> 00:48:26.130
to people normally, reminds me of what X used

00:48:26.130 --> 00:48:29.170
to be like, you know, or Twitter 15 years ago.

00:48:29.719 --> 00:48:32.500
Um, my worry is blue skies that, you know, as

00:48:32.500 --> 00:48:34.079
it grows, they're going to think, how can you

00:48:34.079 --> 00:48:35.579
squeeze some money out of this? And then it becomes

00:48:35.579 --> 00:48:37.940
an algorithmic hellscape. But I think we have

00:48:37.940 --> 00:48:41.159
to kind of, uh, you know, express the value of

00:48:41.159 --> 00:48:43.219
having those spaces where the algorithm doesn't

00:48:43.219 --> 00:48:45.840
dominate the conversation and that you don't

00:48:45.840 --> 00:48:47.880
get reshares just because you're sharing a, you

00:48:47.880 --> 00:48:49.679
know, singing and dancing cat video made with

00:48:49.679 --> 00:48:53.039
AI. Yeah. Well, I mean, I certainly think non

00:48:53.039 --> 00:48:55.340
-algorithmic spaces like WhatsApp or Reddit or

00:48:55.340 --> 00:48:58.320
Discord, you're seeing a growth of those kind

00:48:58.320 --> 00:49:00.780
of spaces because people are sort of fed up with

00:49:00.780 --> 00:49:04.639
the algorithmic stuff. There will always be a

00:49:04.639 --> 00:49:07.340
place, like I said, there's a role for that kind

00:49:07.340 --> 00:49:08.860
of content that makes me feel better when I've

00:49:08.860 --> 00:49:10.780
had a bad day. But there were also times where

00:49:10.780 --> 00:49:14.900
I just want good information in ways that I can

00:49:14.900 --> 00:49:18.409
verify who the person is. Yeah, Rafi agrees.

00:49:18.690 --> 00:49:20.789
They said, I think the overflow of slop will

00:49:20.789 --> 00:49:22.750
give people a renewed interest in controlling

00:49:22.750 --> 00:49:24.909
whether their feeds can be consumed with content

00:49:24.909 --> 00:49:27.449
that is not conducive to a useful flow of information.

00:49:27.869 --> 00:49:29.809
I just wanted to flag as well, above Kara said,

00:49:29.989 --> 00:49:31.869
just wanted to add fact checking can also be

00:49:31.869 --> 00:49:34.130
powerful if used by relatives, friends, act of

00:49:34.130 --> 00:49:36.690
those falling for disinformation. So it's also

00:49:36.690 --> 00:49:39.329
important, you know, to work within your local

00:49:39.329 --> 00:49:42.840
and your family communities as well. It's not

00:49:42.840 --> 00:49:46.500
just online. Yeah, but there's really good research

00:49:46.500 --> 00:49:48.420
about bystander fact -checking, right? So if

00:49:48.420 --> 00:49:51.099
somebody on Facebook says something bonkers and

00:49:51.099 --> 00:49:53.380
you respond, it's unlikely you're going to convince

00:49:53.380 --> 00:49:55.960
the original poster, but you have a lot of people

00:49:55.960 --> 00:49:58.519
lurking who see that and are likely to be like,

00:49:58.519 --> 00:50:00.059
oh, that's interesting. Let me go and read that

00:50:00.059 --> 00:50:02.940
article. So I often say to people, it's not about

00:50:02.940 --> 00:50:05.239
changing the mind of the first person, but recognize

00:50:05.239 --> 00:50:08.559
that you're having benefits in other ways. Absolutely.

00:50:09.079 --> 00:50:11.619
Nick has an interesting question earlier. Could

00:50:11.619 --> 00:50:14.059
governments see epistemic security as defense

00:50:14.059 --> 00:50:16.159
spending, like for example, education reform

00:50:16.159 --> 00:50:18.019
and journalism funding counting towards NATO

00:50:18.019 --> 00:50:20.679
spending targets? What do you think of that idea?

00:50:21.039 --> 00:50:23.579
Has there been any interest from governments

00:50:23.579 --> 00:50:27.059
in general to kind of pursue kind of fact checking

00:50:27.059 --> 00:50:29.980
or verification strategies? Is that being taken

00:50:29.980 --> 00:50:33.000
as seriously as it should be? Yeah, I mean, Elliot

00:50:33.000 --> 00:50:35.199
alluded to this earlier, which is working with

00:50:35.199 --> 00:50:36.780
governments is always really, really tricky.

00:50:36.900 --> 00:50:38.960
Like when we did our project in Brazil, we had

00:50:38.960 --> 00:50:41.980
to basically write a statement for a lot of newsrooms

00:50:41.980 --> 00:50:44.099
to say that we'd never taken money from the American

00:50:44.099 --> 00:50:45.780
government because of the American government's

00:50:45.780 --> 00:50:48.860
role in the military dictatorship. And so we

00:50:48.860 --> 00:50:50.639
never took money from the US government. We never

00:50:50.639 --> 00:50:53.480
took money from the UK government. And so I think

00:50:53.480 --> 00:50:56.199
there's this challenge of like, how can we stay

00:50:56.199 --> 00:50:58.579
independent? But your point is a good one. And

00:50:58.579 --> 00:51:02.059
there are like, for example, many of the Scandinavian

00:51:02.059 --> 00:51:04.019
countries and obviously the countries that the

00:51:04.019 --> 00:51:05.719
nearer you get to Russia, the more they take

00:51:05.719 --> 00:51:08.860
this seriously and see it as a kind of a counter

00:51:08.860 --> 00:51:11.820
disinformation as military spending. The question

00:51:11.820 --> 00:51:14.199
is like how much should they be doing versus

00:51:14.199 --> 00:51:16.639
independent organizations? And I think that's

00:51:16.639 --> 00:51:18.760
the real challenge with all of this. But I certainly

00:51:18.760 --> 00:51:21.940
think that there are many governments who are

00:51:21.940 --> 00:51:23.460
increasingly seeing this as something that they

00:51:23.460 --> 00:51:24.639
should be taking seriously. But I think they

00:51:24.639 --> 00:51:25.960
don't know where to start, right? So there's

00:51:25.960 --> 00:51:27.059
this sense of like, well, we should be doing

00:51:27.059 --> 00:51:28.760
more media literacy, but they don't know what

00:51:28.760 --> 00:51:30.900
format that should take or how to do it effectively.

00:51:31.280 --> 00:51:33.920
And it's very targeted on like school kids as

00:51:33.920 --> 00:51:36.420
opposed to like you need cradle to grave. So

00:51:36.420 --> 00:51:38.860
that's my frustration is that we haven't invested

00:51:38.860 --> 00:51:42.500
properly in innovative media literacy interventions.

00:51:42.780 --> 00:51:44.260
And instead we're like, well, we just would do

00:51:44.260 --> 00:51:46.300
a PowerPoint slide to some 18 year old or, you

00:51:46.300 --> 00:51:48.380
know, 16 year olds who laugh at the teacher who

00:51:48.380 --> 00:51:51.179
hasn't got a clue what Snapchat is. But yeah,

00:51:51.199 --> 00:51:52.619
I'm sure Elliot's got some thoughts on all of

00:51:52.619 --> 00:51:54.980
this. Yeah. And on that point, even regards to

00:51:54.980 --> 00:51:57.340
education, there was a recent government advice

00:51:57.340 --> 00:52:00.219
to curriculum development organisations about

00:52:00.219 --> 00:52:03.679
having more media literacy skills, which really

00:52:03.679 --> 00:52:05.880
for me is critical thinking skills needing to

00:52:05.880 --> 00:52:08.429
be embedded across multiple subjects. You're

00:52:08.429 --> 00:52:10.269
no longer in a society where if you want to influence

00:52:10.269 --> 00:52:12.489
discourse, you have to, you know, have a career

00:52:12.489 --> 00:52:14.469
path to become a journalist or politician. You

00:52:14.469 --> 00:52:16.090
pick up your phone and you're influencing the

00:52:16.090 --> 00:52:18.289
discourse just by looking at a post with five

00:52:18.289 --> 00:52:21.010
seconds longer than another post. We need to

00:52:21.010 --> 00:52:22.329
understand that's the environment we live in.

00:52:22.409 --> 00:52:24.230
We need to educate around that environment. And

00:52:24.230 --> 00:52:25.750
if we fail to do that, then we're just going

00:52:25.750 --> 00:52:27.510
to keep on repeating the same mistakes. It's

00:52:27.510 --> 00:52:30.010
like, um, you know, someone says, Oh, I go into,

00:52:30.010 --> 00:52:32.179
you know, do media literacy skills and we teach

00:52:32.179 --> 00:52:34.480
kids how to fact check headlines, which is things

00:52:34.480 --> 00:52:36.760
that have been said to me. I'm thinking, what

00:52:36.760 --> 00:52:38.739
headline are they seeing to ever fact check?

00:52:38.780 --> 00:52:40.440
Because that's not how they're consuming the

00:52:40.440 --> 00:52:42.360
content they're encountering. You don't have

00:52:42.360 --> 00:52:45.360
this kind of passive consumption of media in

00:52:45.360 --> 00:52:47.460
the kind of 20th century way where you had the

00:52:47.460 --> 00:52:48.920
newspaper that your parents brought home and

00:52:48.920 --> 00:52:53.079
the Scott News was on. now connected to a media

00:52:53.079 --> 00:52:55.159
ecosystem that's algorithmically designed to

00:52:55.159 --> 00:52:57.960
keep you on those platforms. We need to educate

00:52:57.960 --> 00:52:59.820
around that. We can't just say, oh, here's how

00:52:59.820 --> 00:53:01.659
you fact check a bunch of stuff because we need

00:53:01.659 --> 00:53:03.679
to go beyond that. We need to say, actually,

00:53:03.820 --> 00:53:06.260
democracy relies on these things actually working.

00:53:06.760 --> 00:53:09.059
We also need to be able to demonstrate that institutions

00:53:09.059 --> 00:53:11.739
respond to that. We're teaching young people

00:53:11.739 --> 00:53:14.300
that this is how democracy works. This is why

00:53:14.519 --> 00:53:16.940
facts are important and then the government comes

00:53:16.940 --> 00:53:20.820
along and does Iraq in 2003 all over again. That

00:53:20.820 --> 00:53:23.260
just drives the cynicism that sends people to

00:53:23.260 --> 00:53:25.099
these online communities that say, yes, you're

00:53:25.099 --> 00:53:27.159
right to hate the government for being liars.

00:53:27.340 --> 00:53:29.320
Here's lots of information to support that. And

00:53:29.320 --> 00:53:31.579
then in another 20 years where we have enough

00:53:31.579 --> 00:53:33.940
chemical weapons attack in the Middle Eastern

00:53:33.940 --> 00:53:35.440
country, we'll have again people saying, oh,

00:53:35.460 --> 00:53:37.940
well, it's just like it was 20 years ago. I think

00:53:37.940 --> 00:53:42.170
that's the responsibility here. We talk a lot

00:53:42.170 --> 00:53:44.710
about how people have lost trust in institutions,

00:53:45.309 --> 00:53:48.349
but that's often earned. It isn't just because,

00:53:48.349 --> 00:53:50.449
oh, it's a vague idea that they're reading too

00:53:50.449 --> 00:53:52.809
much rushing propaganda or bad information on

00:53:52.809 --> 00:53:54.010
the internet. It's because those institutions

00:53:54.010 --> 00:53:57.170
have failed people in a myriad of ways and those

00:53:57.170 --> 00:53:59.269
institutions need to be able to respond to that.

00:53:59.610 --> 00:54:01.409
But I think they also end up kind of paralysing

00:54:01.409 --> 00:54:03.409
this current information environment because

00:54:03.739 --> 00:54:06.099
What they see is the most salient discourse on

00:54:06.099 --> 00:54:08.820
these platforms that they're engaging with. In

00:54:08.820 --> 00:54:10.980
the UK it's, well, in the US as well, it's immigration

00:54:10.980 --> 00:54:13.579
because it's salient because one side is really

00:54:13.579 --> 00:54:15.739
passionate and hateful, as is the other side,

00:54:16.039 --> 00:54:18.219
against that side. So there's lots of arguing

00:54:18.219 --> 00:54:19.940
online. The algorithm sees, hey, great, loads

00:54:19.940 --> 00:54:22.260
of engagement and starts boosting that. That

00:54:22.260 --> 00:54:24.260
changes the incentives of people on the platforms

00:54:24.260 --> 00:54:26.300
to start posting more stuff about that topic.

00:54:26.599 --> 00:54:28.460
And they have the institutions who exist in those

00:54:28.460 --> 00:54:31.320
same spaces now seeing that as the topic of conversation.

00:54:31.500 --> 00:54:34.199
And they have people like, know, labour chasing

00:54:34.199 --> 00:54:36.280
after reform voters and actually pissing off

00:54:36.280 --> 00:54:39.039
all their own supporters. So we need to understand

00:54:39.039 --> 00:54:41.579
the information environment we're existing in

00:54:41.579 --> 00:54:44.219
even before we start discussing those solutions

00:54:44.219 --> 00:54:46.780
in a really deep way. I'm glad to say that I'm

00:54:46.780 --> 00:54:49.260
consulting with a couple of education establishments

00:54:49.260 --> 00:54:51.119
that are developing curriculum material and they're

00:54:51.119 --> 00:54:52.840
really interested in this kind of deeper way

00:54:52.840 --> 00:54:55.099
of approaching it. So I think we're going to

00:54:55.099 --> 00:54:57.239
get away from the kind of fact checking headlines

00:54:57.239 --> 00:55:00.219
in one lesson that I approach, which I think

00:55:00.219 --> 00:55:01.519
is something that's really, really important

00:55:01.519 --> 00:55:05.019
in this moment. And I think just, I saw somebody

00:55:05.019 --> 00:55:07.159
put into the chat about that blog post from LSE

00:55:07.159 --> 00:55:10.119
and it was from 2018. I remember in lots and

00:55:10.119 --> 00:55:13.239
lots of conferences, 2016, 2017, anybody, somebody

00:55:13.239 --> 00:55:15.000
talked about media literacy, somebody else were

00:55:15.000 --> 00:55:16.980
like, oh, we haven't got the time. We just need

00:55:16.980 --> 00:55:18.780
to get Facebook to tweak its algorithm or we

00:55:18.780 --> 00:55:20.679
just need a spam folder for misinformation. We

00:55:20.679 --> 00:55:23.010
haven't got time. And now we're 10 years on,

00:55:23.130 --> 00:55:24.869
I'm like, what would have happened if we'd actually

00:55:24.869 --> 00:55:27.750
invested in media literacy properly 10 years

00:55:27.750 --> 00:55:29.190
ago, right? Not to say that we wouldn't be where

00:55:29.190 --> 00:55:31.150
we are today, but it wouldn't be quite as bad

00:55:31.150 --> 00:55:33.829
as it is. And I'm with Elliot that we just haven't

00:55:33.829 --> 00:55:36.309
done enough to explain how algorithms shape what

00:55:36.309 --> 00:55:37.750
we see. Like I'm doing a project right now on

00:55:37.750 --> 00:55:40.090
polarization. And when you talk to people like,

00:55:40.329 --> 00:55:42.309
well, you know that we have like polarized candidates

00:55:42.309 --> 00:55:45.070
like Marjorie Taylor Greene or, you know, left

00:55:45.070 --> 00:55:47.710
-wing candidates because the algorithm privileges

00:55:47.710 --> 00:55:50.250
the more extreme you are. Like if you're a nice,

00:55:50.550 --> 00:55:52.630
centrist candidate, you don't get any airtime.

00:55:52.650 --> 00:55:55.050
So that's forced people to become more and more,

00:55:55.050 --> 00:55:57.949
you know. So once you start explaining how algorithms

00:55:57.949 --> 00:56:00.230
have impacted our headlines, the types of politicians

00:56:00.230 --> 00:56:03.659
we see, what politicians say, then I think people

00:56:03.659 --> 00:56:06.179
start to be like, oh, okay, right. So it's algorithms

00:56:06.179 --> 00:56:07.500
are at the heart of so many different things,

00:56:07.519 --> 00:56:09.760
but we've done a really bad job of explaining

00:56:09.760 --> 00:56:11.800
how algorithms work. And most people are like,

00:56:11.900 --> 00:56:13.500
oh, everything I see on Facebook is just the

00:56:13.500 --> 00:56:15.059
people I like. They still don't understand that

00:56:15.059 --> 00:56:16.500
what they see on Facebook is algorithmically

00:56:16.500 --> 00:56:19.179
driven and that they don't see 90 % of the stuff

00:56:19.179 --> 00:56:20.880
that they could. Then people's minds are like,

00:56:20.980 --> 00:56:29.170
ooh, and it's 2026. So yeah, algorithms. I'll

00:56:29.170 --> 00:56:30.909
put all of the links that have been mentioned

00:56:30.909 --> 00:56:32.590
in the chat in the description of the podcast.

00:56:32.730 --> 00:56:34.730
So if you are listening back, you can still check

00:56:34.730 --> 00:56:38.469
them out, including a recent article that Elliot

00:56:38.469 --> 00:56:41.309
wrote for the Voters Institute, which talks about

00:56:41.309 --> 00:56:44.929
how to protect democracies from falsehoods and

00:56:44.929 --> 00:56:47.489
has some examples of how you could teach critical

00:56:47.489 --> 00:56:52.070
thinking. So please do give that a read. Bear

00:56:52.070 --> 00:56:54.590
in mind, we only have about seven minutes left,

00:56:54.610 --> 00:56:58.750
so we have to limit the questions. I wanted to

00:56:58.750 --> 00:57:03.550
go back to the idea of kind of how we talk about

00:57:03.550 --> 00:57:07.349
disinformation. And Caro asked earlier, I remember

00:57:07.349 --> 00:57:09.250
reading Claire that you've become rather critical

00:57:09.250 --> 00:57:11.809
of the term malinformation. Can you elaborate

00:57:11.809 --> 00:57:14.119
on that a little? Are the terms that we're using

00:57:14.119 --> 00:57:16.480
disinformation, misinformation, malinformation,

00:57:16.760 --> 00:57:19.780
fake news, are they confusing, misleading? Are

00:57:19.780 --> 00:57:21.760
they causing more problems than they should?

00:57:24.039 --> 00:57:27.079
Yes, I wrote a piece in the spring of 2023, where

00:57:27.079 --> 00:57:30.179
I basically said that So I wrote a report in

00:57:30.179 --> 00:57:32.960
2017 called Information Disorder, which came

00:57:32.960 --> 00:57:35.860
from a frustration that everybody was using the

00:57:35.860 --> 00:57:37.960
term fake news, which had been weaponized and

00:57:37.960 --> 00:57:39.420
actually didn't describe a lot of the problems.

00:57:39.480 --> 00:57:41.159
A lot of this content isn't fake and it certainly

00:57:41.159 --> 00:57:43.500
doesn't look like news. So I wrote this report

00:57:43.500 --> 00:57:45.559
and it had like it was a very long report, but

00:57:45.559 --> 00:57:47.960
it had this Venn diagram in around misinformation,

00:57:48.059 --> 00:57:50.360
disinformation, malinformation. And that Venn

00:57:50.360 --> 00:57:52.579
diagram has appeared on a lot of slide decks

00:57:52.579 --> 00:57:54.280
over the years. Like I've seen them pop up and

00:57:54.280 --> 00:57:56.800
I'm like, oh, that slide deck. But I sort of,

00:57:56.980 --> 00:57:59.300
I wrote this piece to say that wasn't helpful,

00:57:59.300 --> 00:58:01.820
right? Because we got obsessed with trying to

00:58:01.820 --> 00:58:03.940
define different examples. Like, but what bucket

00:58:03.940 --> 00:58:05.820
do we put it in? Right. And so much of this was

00:58:05.820 --> 00:58:08.139
about intent and we couldn't actually tell. And

00:58:08.139 --> 00:58:09.980
so this piece that I wrote was basically saying

00:58:09.980 --> 00:58:12.739
we, by doing that, we got fixated on thinking

00:58:12.739 --> 00:58:15.000
that we could take the complexity of the ecosystem

00:58:15.000 --> 00:58:17.920
and make it simple. But, and it meant that we

00:58:17.920 --> 00:58:19.579
completely forgot to look at a whole host of

00:58:19.579 --> 00:58:22.360
different factors, right? And so I talk about

00:58:22.360 --> 00:58:24.119
missing disinformation, right? Like, so if I'm

00:58:24.119 --> 00:58:26.880
in an anti -vax community and there's an image

00:58:26.880 --> 00:58:29.300
I have from Instagram with a woman who's holding

00:58:29.300 --> 00:58:31.519
a binder full of scientific evidence, right?

00:58:31.820 --> 00:58:33.800
She's reading the evidence. She's not somebody

00:58:33.800 --> 00:58:35.699
who just like forwards something in the coffee

00:58:35.699 --> 00:58:37.820
shop without checking. She's looking at this

00:58:37.820 --> 00:58:39.940
stuff all day long, but she's just reached different

00:58:39.940 --> 00:58:42.260
conclusions to me. Is she a disinformer or is

00:58:42.260 --> 00:58:44.039
she a misinformer? She's neither of those things,

00:58:44.340 --> 00:58:47.019
right? The people who stormed the Capitol. So

00:58:47.019 --> 00:58:49.800
the more I was doing this work, I was like, none

00:58:49.800 --> 00:58:51.639
of those early definitions make sense when you

00:58:51.639 --> 00:58:53.780
start operationalizing it. Or the people that

00:58:53.780 --> 00:58:55.900
use the, when Musk changed the blue tick thing

00:58:55.900 --> 00:58:58.659
to say it was Eli Lilly and said, now everybody

00:58:58.659 --> 00:59:03.480
can get like free. insulin. What was that? Was

00:59:03.480 --> 00:59:05.000
that misinformation or disinformation? Was it

00:59:05.000 --> 00:59:06.920
causing harm? Well, it caused harm to some very

00:59:06.920 --> 00:59:09.400
rich fat cats on Wall Street, but it was helping

00:59:09.400 --> 00:59:12.880
the public interest. So I think that these definitions,

00:59:13.239 --> 00:59:16.079
maybe they were necessary in 2016 to take a complex

00:59:16.079 --> 00:59:18.539
problem and make it make sense. But I actually

00:59:18.539 --> 00:59:20.920
really try and avoid a lot of the terms now,

00:59:20.980 --> 00:59:23.539
partly because they've been weaponized. But secondly,

00:59:23.579 --> 00:59:25.440
I don't think any of this is as straightforward

00:59:25.440 --> 00:59:29.579
as we like to think. And I think by doing so,

00:59:29.699 --> 00:59:31.090
it has prevented us from really understanding

00:59:31.090 --> 00:59:33.889
some of the key drivers to these problems. And

00:59:33.889 --> 00:59:35.769
so for that reason, I've changed my language.

00:59:36.710 --> 00:59:40.670
So yeah. And I do use those terms in my discourse

00:59:40.670 --> 00:59:43.090
framework, but that's because I can then build.

00:59:43.349 --> 00:59:46.190
So I use those as the elements and then I explain

00:59:46.190 --> 00:59:48.010
what the drivers are, who the people are, you

00:59:48.010 --> 00:59:50.769
know, the types of people. groups or people who

00:59:50.769 --> 00:59:52.929
are spreading this stuff and the process behind

00:59:52.929 --> 00:59:55.050
that. So that's, I think, the thing that's often

00:59:55.050 --> 00:59:56.929
overlooked. It has to be about when we're talking

00:59:56.929 --> 00:59:59.289
about this information, about the discourse,

00:59:59.530 --> 01:00:02.110
not just about, you know, this is bad information

01:00:02.110 --> 01:00:03.630
and it's getting out to people and therefore

01:00:03.630 --> 01:00:05.670
we have to make the information. It's like...

01:00:06.400 --> 01:00:09.059
Why does this discourse emerge in the first place?

01:00:09.619 --> 01:00:11.880
Because that lets us off far too lightly and

01:00:11.880 --> 01:00:14.599
it lets off institutions far too lightly. Because

01:00:14.599 --> 01:00:16.579
again, it comes back to this idea old people

01:00:16.579 --> 01:00:18.099
have lost trust in institutions, but we need

01:00:18.099 --> 01:00:20.480
to understand why they've lost trust in institutions.

01:00:21.159 --> 01:00:23.320
What was that trust even based on in the first

01:00:23.320 --> 01:00:25.880
place? We don't really ever question that. Um,

01:00:25.880 --> 01:00:28.360
and if we don't do that, then we'll never actually

01:00:28.360 --> 01:00:30.639
be able to find workable solutions. And I think

01:00:30.639 --> 01:00:33.579
that's where, you know, a lot of post 2016 counter

01:00:33.579 --> 01:00:35.619
disinformation efforts did fail because it was

01:00:35.619 --> 01:00:38.039
so much about the information rather than the

01:00:38.039 --> 01:00:39.960
discourse that was happening and the role of

01:00:39.960 --> 01:00:42.019
institutions in being responsible for that discourse

01:00:42.019 --> 01:00:46.860
in the first place. Yeah. And I think just to

01:00:46.860 --> 01:00:49.219
jump in, the question actually was about malinformation,

01:00:49.219 --> 01:00:51.940
which was this term that I coined with my co

01:00:51.940 --> 01:00:54.139
-author, Hossein Darikshan. And the idea was

01:00:54.139 --> 01:00:57.940
when look, genuine information is used to cause

01:00:57.940 --> 01:01:00.440
harm, right? And so I've done work with the ICRC

01:01:00.440 --> 01:01:03.860
and the UN, like some very effective malinformation

01:01:03.860 --> 01:01:06.099
is when you go back and talk about the ICRC's

01:01:06.099 --> 01:01:08.719
role during World War Two, right? Like, or you

01:01:08.719 --> 01:01:10.480
talk about the UN and what some of the peacekeepers

01:01:10.480 --> 01:01:13.119
have done, right? Like that is not false information.

01:01:13.420 --> 01:01:15.760
It's like that is it's being is the weaponization

01:01:15.760 --> 01:01:18.300
part that is useful and interesting. labeling

01:01:18.300 --> 01:01:20.219
it isn't that helpful, right? And so I think,

01:01:20.480 --> 01:01:22.179
but what I was trying to make the point by talking

01:01:22.179 --> 01:01:23.760
about malinformation is that everybody was so

01:01:23.760 --> 01:01:25.739
obsessed with false information, you could only

01:01:25.739 --> 01:01:27.739
flag false information without realizing that

01:01:27.739 --> 01:01:30.960
the smartest actors were weaponizing genuine

01:01:30.960 --> 01:01:33.260
information because you couldn't go to a platform

01:01:33.260 --> 01:01:34.880
and be like, you want to take that down, it's

01:01:34.880 --> 01:01:36.760
disinformation because it's not, it's genuine

01:01:36.760 --> 01:01:39.019
information, right? And so again, to Elliot's

01:01:39.019 --> 01:01:40.880
point, so much of this is understanding historical

01:01:40.880 --> 01:01:44.059
context, understanding institutions properly

01:01:44.059 --> 01:01:46.150
and the discourse. around them to understand

01:01:46.150 --> 01:01:47.650
how they're being weaponized. It's not about

01:01:47.650 --> 01:01:49.530
is it true or false. And that's, I think, where

01:01:49.530 --> 01:01:53.030
we got lost a little bit. Thanks for giving that

01:01:53.030 --> 01:01:55.150
extra context. We're going to have to wrap in

01:01:55.150 --> 01:01:56.829
a second, but I wanted to just read out this

01:01:56.829 --> 01:01:59.329
point. One of somebody in the chat said, 30 years

01:01:59.329 --> 01:02:02.250
ago, my views aligned with today's Marga. Many

01:02:02.250 --> 01:02:04.710
of my friends from those days are Maganal. It

01:02:04.710 --> 01:02:07.309
was people sharing facts as seeds, not demands

01:02:07.309 --> 01:02:09.929
that helped me change, that led me to be the

01:02:09.929 --> 01:02:13.550
centre -left pro -facts person I am now. So really

01:02:13.550 --> 01:02:17.329
builds on a lot of what you were saying. Just

01:02:17.329 --> 01:02:19.369
to wrap up, a lot of people have been asking

01:02:19.369 --> 01:02:22.630
how they can get in contact with you both if

01:02:22.630 --> 01:02:26.710
they have follow up questions or want to highlight

01:02:26.710 --> 01:02:29.630
amazing things that they're working on to you.

01:02:30.210 --> 01:02:34.510
Where's the best way to contact you? I'm going

01:02:34.510 --> 01:02:38.230
to put my email in the chat. So I might not be

01:02:38.230 --> 01:02:39.809
able to get back immediately, but I always love

01:02:39.809 --> 01:02:41.789
to hear people and what they're doing. And so

01:02:41.789 --> 01:02:44.409
it would be my pleasure to get into a conversation

01:02:44.409 --> 01:02:47.610
with anybody if they want to continue it. Amazing.

01:02:47.690 --> 01:02:49.730
You can contact me directly at my email as well.

01:02:50.539 --> 01:02:53.000
Fabulous. Thank you so much both for your time

01:02:53.000 --> 01:02:56.119
today. And thank you to everybody who's joined

01:02:56.119 --> 01:02:58.099
the audience as well. Thank you for your fantastic

01:02:58.099 --> 01:03:00.679
questions. And we'll be back in two weeks time

01:03:00.679 --> 01:03:03.539
with another stage talk. But for now, this recording

01:03:03.539 --> 01:03:05.699
will be up hopefully by tomorrow for you all

01:03:05.699 --> 01:03:08.960
to listen back to. And yeah, again, Claire, Elliot,

01:03:09.039 --> 01:03:10.800
thank you so much for coming in and chatting

01:03:10.800 --> 01:03:13.960
with us. Thank you so much. Thank you. I loved

01:03:13.960 --> 01:03:17.300
it. Thank you for listening to the stage talk.

01:03:17.610 --> 01:03:20.730
If you'd like to catch a stage tour live where

01:03:20.730 --> 01:03:23.849
you can ask the guest questions, join the Bellingcat

01:03:23.849 --> 01:03:30.250
Discord server by visiting www .discord .gg slash

01:03:30.250 --> 01:03:33.150
Bellingcat. The music you've heard is titled

01:03:33.150 --> 01:03:37.070
Dawn by Newer Self and is courtesy of Artlist.