Subscribe
Share
Share
Embed
The Honest Take is HonestReporting's long-form interview series bringing together leading voices to examine how Israel and the Middle East are covered — and what goes into creating and feeding the narratives that target Israel.
Each episode features leading experts, analysts, researchers, and journalists who work on media bias, terrorism, NGO accountability, foreign influence, antisemitism, and international institutions. These are professionals directly involved in investigating how narratives are shaped, amplified, and protected across global newsrooms.
The conversations go beyond breaking news to unpack:
• Media bias and misinformation about Israel
• How terrorist organizations exploit humanitarian and civil society frameworks
• The role of NGOs and international bodies in shaping public perception
• Foreign state influence on Western media and education
• Why context disappears in reporting on Israel and the Palestinians
This series is for listeners seeking in-depth analysis, evidence-based discussion, and accountability in journalism on Israel and the Middle East.
Honest Take Ashley Rindsberg X Audio.txt
English (US)
00:00:00.000 — 00:00:20.600 · Speaker 1
We all kind of understand that we're in an information warfare that's going on all the time, all around us. A lot of people out there might be thinking to themselves, I don't go to Wikipedia. The reality is that Wikipedia comes to you. It comes to you in the form of AI, which trains on its data. There's so easily to hijack the processes on the site that they're basically no longer function.
00:00:22.480 — 00:01:23.500 · Speaker 2
Welcome to the honest take. For years we've called out media bias against Israel. But today, the most powerful narrative engine isn't a newsroom. It's Wikipedia, it's Reddit, it's Google, it's AI. Wikipedia fills Google's knowledge panels, Reddit threads, gets scraped into search results, and both are fed directly into the systems that answer questions for millions of people every single day.
So when those platforms shift, the internet's baseline understanding of Israel shifts with them. Ashley Rensburg has been investigating how that happens. He's the founder of Neutral Point of View, and he's an investigative journalist and an author who has exposed coordinated editing campaigns around Israel, the internal power structures inside Wikipedia, and how Reddit content is amplified upstream before it's repeated back to the world as fact.
Even if you aren't on Reddit or regularly refer to Wikipedia, you're still an unsuspecting participant in today's information warfare.
00:01:23.540 — 00:01:37.500 · Speaker 3
Ashley, thanks so much for joining us. You've spent years looking at this Wikipedia influence, Reddit influence. Where did you see this divide opening up between like the the actual neutrality and, you know, as you call it, poisoning?
00:01:37.540 — 00:03:18.100 · Speaker 1
So I was I was doing some work in and around tech, media and tech for Pirate Wires, where I was senior editor, and I came across this really interesting study by a friend and colleague of mine named David Rosado, who is a machine learning researcher. So he looks sort of at bias on AI systems. He's been doing this for a very long time, and he's amazing at it.
So he produced this study showing that there is pervasive bias on Wikipedia as kind of a political bias. That was pretty clear to my reading. So it showed that if you take a political figure in the United States and he looked at senators, members of the House, presidents, Supreme Court justices, he also looked at journalists.
You would see that if they were on the right, they were more likely to be associated with negative sentiment. So sort of bad sounding terms on Wikipedia. And if you looked at figures on the left, the opposite was true. So they would be more likely, they are more likely to be associated with positive sentiment.
And he did this across a number of different sort of categories. He looked at U.K. British prime ministers like it just it just kind of was there and showed up across the board. It's not an isolated case in any one group of people. And I started to think to myself, like, if we're seeing this show up in Wikipedia, which is supposed to be resistant to bias or to minimize bias.
At least there's got to be a lot more to the story that we're not seeing. The question of how did it get to this point? What are the drivers of the bias? How does the how does this work and what is the impact of it? And that's where I started really digging into Wikipedia first and trying to understand what's actually going on when you peel away the outer layer.
00:03:18.140 — 00:03:37.220 · Speaker 3
And so like a lot of people assume that there is some bias in Wikipedia. At least that's how it's been for me since it started. But that if it exists, it's it's random or accidental or just uninformed people editing from what you found in your research, it's much deeper than that. There's more coordinated efforts.
00:03:37.340 — 00:06:30.570 · Speaker 1
There's so Wikipedia, it's sort of helpful to understand it as two big pieces. And the one big piece is the website that we all know and some of us love or used to love. Um, and then aside from the website, which is supposed to operate, at least the contents are supposed to operate independently by volunteers.
Aside from the website, there's Wikimedia Foundation. That's the NGO that owns Wikipedia. And that's a foundation that was started early days to be able to take money in donations. It has about a $200 million budget, and it's really quite a behemoth in the world of NGOs. So when you think about these two different pieces, they both interact with each other and both of them have a significant issue.
So on the platform itself, what we see is that it is just a hot zone of coordinated editing of various kinds, and that can be ideological. And where maybe these are more or less they're not good faith because they're manipulating the system, but they believe in what they're doing. Then you can have people who are doing this for money.
So they are just working on behalf of brands, companies, individuals, because people understand Wikipedia is by far one of the most valuable tools for SEO. Search engine optimization that exists today. So they want a piece of the action. And it's, you know, pound for pound. It's not that expensive to do it.
So a lot of people do it. And there's a lot of agencies that are doing this kind of thing. And then you have the more nefarious stuff, which is um, it sort of blurs between the ideological stuff, but it's sort of state aligned or state backed, or they're indications that there is some external body that is coordinating this stuff beyond just a few ideologically charged editors who happen to agree with each other.
And that's kind of organic. So you have this spectrum of coordination on Wikipedia that's absolutely there. We obviously we'll talk a little bit more about how that shows up specifically. But then on the Wikimedia Foundation side, you have it their own kind of bias. And that's an ideological commitment.
It was always kind of baked in, as you kind of mentioned. It was always kind of there because of the, the geography. The the foundation was started mostly sort of in San Francisco. They actually were founded in in Florida, but they quickly moved to San Francisco. You attract a certain kind of crowd. There's a culture around that crowd.
I think 95% of those people were were actually very well intentioned. They really wanted to build something good. But over time, the organization becomes an institution. It has its own institutional ethos. And then something really flipped in 2016, 2017, where Wikimedia Foundation doubled down on its what used to be sort of a center left liberal bias, and they turned it into something much more solidly on the left, the ideological left.
00:06:30.770 — 00:06:34.370 · Speaker 3
Israel is a recent case study of your work. Wikipedia calls it.
00:06:34.370 — 00:06:38.050 · Speaker 1
The IPA stands for Palestine, Israel. Articles.
00:06:38.090 — 00:06:42.850 · Speaker 3
Yeah, exactly. And that that is one of the more contentious areas.
00:06:42.850 — 00:08:17.570 · Speaker 1
I would probably venture to say it's the most at this point. Maybe there's American politics, there's a lot of activity there. But in terms of, um, the intensity over the, the surface area of, of that space, it's got to be at least one of the top 2 or 3 areas where the level of contentiousness it's actually contentiousness at this point is a euphemism.
It's an ideological war. Um, I've reported and investigated what I call the gang of 40, which is about 40 ideologically committed editors who edit almost entirely in Pia in the Palestinian-Israeli articles space. They're coordinated with each other, um, whether or not they coordinate on a, on a separate platform in addition to Wikipedia, I don't know.
But their editing shows a pattern of de facto coordination, and they are extremely anti-Israel. Very, I would say at minimum pro Palestine. Some of them are pro Hamas, some of them are openly pro Hezbollah. Uh, many of them, and they are very sophisticated. They have a very clear, close understanding of how to use Wikipedia on its tools and its rules and guidelines and all this kind of thing to their benefit.
And they also have numbers. So they have quite a lot of people. We have three dozen plus or more. At this point, people who are really dedicated to editing in this space and they're spending many hours a week, some of them are spending upwards of a dozen plus per week and times 40 over the course of years. And what you end up with is a complete capture of that entire topic space.
00:08:17.730 — 00:08:28.290 · Speaker 3
Do we know if that's ideologically driven or like, are these just people who are passionate about the subject, or is there some deeper coordination funding?
00:08:29.050 — 00:11:15.510 · Speaker 1
Um, whether or not they are being sort of run as an operation by some third party? I don't know. I don't have evidence of that. My, in my sense is that that's not the case here in this particular with this particular group, there might be others, but with these guys, they are deeply, ideologically committed for whatever reason.
I think a lot of them, a number of them are academics, sort of Anglo academics. Very few of them are, um, Palestinian or Arab. They mostly seem to be the typical Wikipedia avatar, which is middle aged white men. Um, so that seems to be who they are. But their activities are extremely, extremely adversarial, antagonistic.
They do things like, uh, one of them. Iskander three. Two, three. He'll remove just wholesale mentions of Hamas terror attacks, or he'll move, remove like dozens and dozens of mentions of human rights abuses and crimes committed by the Islamic Republic of Iran. Another one has a user box that's pledges his support to Hezbollah, and he edits accordingly.
And then some of the stuff is very subtle, where they will do little things, like they'll change the name of a historic term. So, for example, there was an article titled Hellenistic Judea and They they changed into Hellenistic Palestine, which is a it's an anachronism. The Palestine was was only a term used much later in the history.
But it's these little shifts or taking the Dead Sea Scrolls. A photo of the Dead Sea Scrolls out of the Israel article. The article on Israel removing that photo to make these little tiny shifts that shape the topography. So they're severing historical ties between Israel and the Jewish people. They're whitewashing the crimes of Hamas and Hezbollah and Iran in one cases.
In one case, that same editor I mentioned, Iskander. He would he really put a lot of focus on removing mentions of Hamas's 1988 genocidal charter. So things that indicate clear intent. You know, there's not there's not like, oh, I really don't you know, I don't particularly agree with your view on Tom Segev book about the 67 war or something like that.
It's not it's not that. It's these are things that are moving in a certain pattern direction. And there's also a clear pattern of editor coordination. They work in small clusters in order to evade detection by Wikipedia admin. So these are things that editors would not naturally and organically do with each other, just statistically speaking.
So between all these things, you get a picture of a group that is ideologically committed, coordinated among itself and having tremendous impact.
00:11:15.550 — 00:11:34.790 · Speaker 3
I guess we can set the stage a little bit. Wikipedia is not supposed to have that bias, and there's not supposed to be coordinated efforts. You just mentioned. The way they're working seems to be intentionally to skirt the rules and regulations there. How does that I mean, how is Wikipedia supposed to work?
00:11:34.830 — 00:15:18.390 · Speaker 1
I mean, Wikipedia is a is, I would say, a product of a lot of idealism from the early internet. So the idea was that if you bring enough people into a given topic and have them all working in good faith and adhering to certain principles and rules, that the the sort of the distribution of different viewpoints would level out the the noise, the false ideas, the false claims, and what would remain is mostly the signal, the true claims that I think may have worked, and I think it possibly did work for a good few years, ten years, 10 to 15 years.
We're in a different information environment today where every single person who's a part of it, whether you're consuming it or operating in it or investigating it, like me, we all kind of understand that we're in an information warfare that's going on all the time, all around us, and Wikipedia being what it is in terms of its importance in the information ecosystem.
So, you know, we think a lot of people out there might be thinking to themselves, I don't go to Wikipedia. The reality is that Wikipedia comes to you and comes to you in the form of AI, which trains on its data comes to you in the form of Google, which relies disproportionately on Wikipedia for almost every single topic search you do.
So the first result you're often going to get. Most times you're going to get is a Wikipedia article on your search, and there's also a knowledge panel that gets filled out by Wikipedia. So in this kind of environment where we are all fighting for ideological real estate, there was no chance that Wikipedia would be able to to survive its former, um, the spirit of the site that was that it was built on and sort of these innocent, naive set of rules and guidelines and principles.
There's so easily to manipulate them and to hijack the processes on the site that they're basically no longer function. So, you know, the short answer to the question is yes, you are not supposed to do coordinated editing. You're not supposed to do you're not supposed to edit on just a single topic and that be your thing?
I'm only editing. It's called a single purpose account. That's not. It's not something you're supposed to do. You're not supposed to be pushing a point of view. So there's a a very famous policy on Wikipedia called POV, which is neutral point of view. You're supposed to always have a neutral point of view on a given topic, which at this point in time is almost kind of a laughable idea on Wikipedia.
That's just like in anything that is meaningful, it just does not exist at all. Um, a number of other types of, of policies like this. But at the end of the day, Wikipedia has become something that it was never intended to become or let's say, the mechanisms that run it, which initially was about consensus and consensus.
When you when you speak with Larry Sanger, who was the co-founder, is the co-founder. He'll talk about consensus in the early days is like, let's represent every significant viewpoint on the topic. Every mainstream accepted viewpoint. And if there's a fringe viewpoint, you can sort of label it as a fringe viewpoint or not include it today.
Wikipedia works by sort of head count, so you'll have a bunch of editors arguing on a topic, and some admin will come in and he'll say, one, two, three, four, five, six of you believe it's a and only four of you believe it's B. So we're going with the six. So whoever the six happened to be often will win the point.
They will win the debate on that basis. But that's actually not consensus. That's just your your voting. It's brute force. Like you can force things to go the way you want them to. If you have the manpower, which in often cases these groups do.
00:15:19.110 — 00:15:28.030 · Speaker 3
Who polices this? Who? How does Wikipedia prevent this or at least attempt to prevent this? Is there even an apparatus that works to do that?
00:15:28.110 — 00:17:28.560 · Speaker 1
So the the primary mechanism are the admins who are also volunteers, and they are supposed to, you know, make these sort of like referees, like they're sort of reffing a game, and if there's a call here or there, they make the call, then they're on the ground. Um, there's not that many of them that are active.
So we're talking about 7 million articles on the site, and let's say there's around 800 admins, uh, 600 maybe are really active. And when we're thinking about active, we're not talking about like, this is my 9 to 5. Like if you go on to a different site where they have moderators who are paid, that's their job.
These are people who are doing this in their spare time. Might be an hour. It might be a couple hours. Uh, then you have above them, you have sort of a couple other layers of of more technical admins. They're called bureaucrats who are can make admins or, or D-list admins. Um, and then sort of above that group, you have arbitrators who are the quote unquote Supreme Court of Wikipedia.
There's 13 of them. Um, I don't think they're all active right now. I think maybe 10 or 11. So again, we're talking about ten people for 7 million articles and and granted, not all those articles are active and not all of them are demand their attention. But even if we say like a tiny fraction of them, like 20 or 30,000 do demand their attention.
It's just it's literally physically impossible for them to be able to deal with that kind of volume. So the the ways that the site is supposed to police this kind of thing. Um, it doesn't it doesn't actually work today, which is evidenced by the fact that all these things carry on without anyone stopping them.
Uh, Wikimedia Foundation itself does. From what I can see, virtually nothing. Like they have no investigative tools. They don't provide investigative tools. They provide some basic like metrics on on Wikipedia itself, which can be helpful, but they have not evolved their toolset to be able to look for this kind of thing or to identify it.
They show no interest in doing that.
00:17:28.560 — 00:17:29.439 · Speaker 3
And
00:17:30.920 — 00:17:31.920 · Speaker 3
do they even know about it?
00:17:32.040 — 00:18:34.740 · Speaker 1
Largely, no. Um, they not only do they not know about these things, but there are there are rare instances they once I think it was in 2021. They they identified a group of pro-China editors who were like running this basically coordinated operation. There was 300 of them in the group. They banned six of them, but they only banned the six because they alleged that these six people were threatening physical violence against their their adversaries on Wikipedia, which were pro-democracy Hong Kong people.
So it got to the point where, like, okay, if there's going to be actual physical violence, like someone's going to get killed, then we're going to ban a user. But up until that point, they did nothing. And since that point, they've done very, very little. There's Wikimedia Foundation not doing not really doing that for whatever reason.
I don't know what it is. And um, and then there's like, you know, two journalists, maybe me and another guy, Aaron Bandler, um, maybe 1 or 2 more, but but very, very few.
00:18:34.740 — 00:18:53.960 · Speaker 3
And it's not the kind of thing where somebody who's has a more balanced point of view could. Like, for instance, I've maybe edited 1 or 2 edits on Wikipedia in my entire life. It's not as easy as me. And, you know, a bunch of other people just jumping in and undoing these edits.
00:18:54.000 — 00:20:04.420 · Speaker 1
No, you will you will not succeed in that, unfortunately. Um, Wikipedia. There's a steep learning curve. It's almost like, you know, I don't know if anyone remembers. There's that movie from the early 2000. Ocean's 12 is the Ocean's 11 sequel, and, uh, Vince Costello is the sort of cat burglar. He's got to go through this room of crisscrossing lasers in order to steal the thing, in order to be able to get to a point where you can edit contentious topics, uh, effectively, you need to be able to, like, maneuver between the lasers, not make any missteps, like, not do this or that where you'll you'll get blocked from a topic.
You've got block from the entire topic area because you just didn't know that you're supposed to do this thing, even to get to the point where you can edit within Palestine Israel articles, you need to have 500 confirmed edits. And so this already is a high bar for most people will take them. I would imagine if they're, you know, just an average person, it would take a year or so to learn the system, practice, make, you know, get, make the mistakes, try to figure it out and you get to the point.
Then you're confirmed editor. You're you're you get your 500 edits and
00:20:05.460 — 00:20:43.200 · Speaker 1
you will try to edit something that you think is is correct in your view. And 2 or 3 of the guys that have been talking about from the gang of 40 will will swarm in. They'll immediately note that this is going on and they will these these are old hands. You know, they've been doing this for for ten years, 15 years.
Some of them, they know exactly what to do. They know exactly which moves to make. They know how to outlast you, exhaust you, to trip you up. Um, the the chances that you're actually going to be able to make significant substantive edits over time, unless you're spending literal years on this, are extremely low.
00:20:43.240 — 00:21:13.440 · Speaker 3
We have this program called, uh, Scroll Smarter That, which is an educational series on media literacy in the online age. Um, since everyone's looking at their phones now, one of the examples that we point out is the Zionism page, which seems to be almost ground zero of this fight. If I'm not wrong, that's now been locked.
The definition of Zionism being to remove as many Arabs as possible from the land of Israel.
00:21:13.480 — 00:25:09.160 · Speaker 1
Yeah. So the guys I've been talking about, the gang of 40, a number of them worked together to implement that definition and to sort of get to a point where the Zionism article, which most likely at this point is the single most important resource defining Zionism to the online world, they equate Zionism with, um, ethnic cleansing.
Basically, if we're to use a term which, you know, not a surprising move on their part at all, that's this is like one of their core core objectives. They have key objectives and core objectives that you can see. That's absolutely one of them. But the thing that they did next was quite shocking, because after my initial report came out about the gang of 40 in October 2024, arbitration committee was took notice or alerted or whatever happened there and instituted a what's called Pia five, which is an investigation into the gang of 40 and coordinated editing in the space.
A number of the leaders of the gang of 40 ended up getting topic banned. They could not they could not edit in that space, um, for at least a year when they can appeal. And in that time, just before that, those bands were about to be handed down. They pull off this maneuver and the Zionism page, where they changed the definition of Zionism to make it equated with ethnic cleansing.
And then they do this thing that I personally, and no one I know has ever heard of in this context, which is they take this concept of a moratorium on Wikipedia and they use it. They implement it to lock the opening section, the lead section of the article, which is the most important by far. It's what most people, 90% of people are just going to read the lead.
And it's also what gets ingested into most of the AI and other downstream platforms. They freeze it in place for one year, which is not coincidentally the length of their the minimum length of their topic ban. So until they become unbanned, nobody else can touch it so they can come back and defend it and defend what they did there.
Um, effectively. So when you think about this, if like, you zoom out and you think about this in terms of like, uh, this kind of high concept battlefield maneuvering. It's extremely brilliant. You know what I mean? It's tactically brilliant. Like, they are creative. They are dedicated. They work together in this team.
And these, like, neat little formations. And the outcomes speak for themselves. Like they you now have Zionism defined online by a group of people who are just about openly dedicated to tearing down Zionism and freezing their definition in in place and nobody objecting to this fact, nobody rolling it back on Wikipedia.
Nobody's saying maybe this shouldn't be done in this way. Nobody's stepping in. Except I will say, actually, it was a slightly separate topic, but Jimmy Wales, co-founder, Wikipedia, tried for a moment to step into the Gaza genocide article, which had also been captured by the same group. He too was just.
He was defeated. I mean, we're talking about a man who, until quite recently had veto power over every single thing that happened on the site was shut down, cut down? He was accused of self-dealing by doing this, by just writing on the talk page of the site where people go to discuss the edits made to an article, and then they refer them to arbitration committee for a number of violations.
So he basically got dragged into wiki court for, for just trying to, um, to, to, let's say, allow cooler heads to prevail. So it gives you a sense of how broken the system is when even its co-founder and by far the most important figurehead, cannot make any headway.
00:25:09.640 — 00:25:29.340 · Speaker 3
That's wild. He's created. It's it's like it's Frankenstein in it. That's even a word, you know, come back to bite him. Um, so, yeah, the gossip genocide. I'm interested. Now, what was the, um, the specifics with that? Because obviously that is one of the more contentious words being thrown out there.
00:25:30.060 — 00:28:00.910 · Speaker 1
So the idea here is that Wikipedia can talk in sort of two different kinds of voices, and this is kind of very natural. This makes, you know, it's intuitively natural where you say, many people say that Israel committed genocide is a different statement from saying Israel committed genocide. So if you're an activist and say Israel committed genocide, well, you're doing your job.
If you're an encyclopedia and saying Israel committed genocide flat out in the same way that you would say the sky is blue, or, um, Donald Trump is the president. That's a very different kind of claim. So the second kind of claim, when you're when you're just writing it out without citing who said this or makes this claim, they call it wiki voice.
You're speaking in Wikipedia's own authoritative third party, quote unquote, neutral voice. And that's basically what the, the, the broad strokes of what they did in the Gaza genocide article, which is to say, we're going to flip this from saying a number of groups or various parties or critics or whoever allege Israel committed genocide in the war to Israel, committed genocide in the war.
Now, Wikipedia says this and makes this claim where it's a whatever you want to say about it. It's a contested claim. Whether you agree with it or deny it, it's not something that's established. I don't know that it would ever be established in real time when we're talking about a war. Um, you know, there was the famous or the some people would say the infamous ICC case, and I don't remember if the ICC or the ICJ or the president came out later, went to the BBC and said, we didn't actually find that Israel was possibly committing genocide.
What they found was Israel, that the Palestinians had a plausible right to bring a case of genocide, and there was no finding. So we have a we have the situation where it's a it's open ended but not according to Wikipedia. And the reason that it's not according to Wikipedia is not because, you know, whatever number of users, 100,000 users voted or something like that, or they they selected, like the 20 most important editors who are experts on the Middle East and Israel, it's just because enough members of the gang of 40 worked on this hard enough for long enough, and in a way that was sophisticated enough to get it done.
And that's it's as simple as that.
00:28:01.030 — 00:28:20.150 · Speaker 3
This isn't just Wikipedia to you have done research into Reddit, which is another, maybe the biggest driver of AI information from, you know, from which AI scrapes is are those the same people? Is that a similar? Um, how does that work?
00:28:20.190 — 00:32:01.710 · Speaker 1
That is a quite a separate kind of operation on Reddit. It has a lot of the same contours, but I think the difference being that on Wikipedia, a lot of the gang of 40 people, they're real Wikipedians. That's my sense. At least I don't. You know, they they know that the stuff ends up downstream. They know it's important.
But I think they would do this anyway. Even if even if Wikipedia did not feed Google and AI, they they find it so important that they would just do it because this is what they do on Reddit. I think it's a different kind of operation. I think it's much more, um, I think there's much more external influence on the operation and much higher degree of mindfulness by that operation of the intent to affect what we call downstream platforms and to to do this kind of data poisoning.
So what we're seeing on Reddit is this group mostly it's sort of base camp. We can say like it's sort of located out of a community. So Reddit, the difference between Reddit and Wikipedia. Wikipedia is organized into articles. That's the primary output. If you think about Facebook, it's like the way that it's organized is you and your friends.
That's your own little community of little network. Reddit works differently. Reddit works. It's organized by communities that are topic based or interest based. So there could be an if there is a community. We call them subreddits on Reddit for Israel. There's one for Jews, there's one for Palestinians, there's one for your favorite soccer football team, whatever.
For thousands and thousands of topics like that. Thousands of subreddits. In this case, the subreddit dedicated to Palestine is called r Palestine. Subreddit Palestine is where the the locus of the activity happens, and what the operation is doing, among other things, is very, very active. What they primarily are doing is taking material straight from the channels belonging to terror organizations the Hamas, Hezbollah, Houthis.
About a dozen others. They translate it. They pull it from telegram into Reddit, and then they spread it from Reddit and then onto other platforms as well. And they organize themselves on discord, which is a messaging app, sort of like Slack or Teams, but it's just kind of independent. They run command and control on that discord, and they coordinate a lot of content, a lot of posting.
They coordinate so that when somebody from their network does a post that they find important, everyone goes and likes it. So it it bumps it up through the algorithm. This is just a quirk of Reddit that you're able to do this the way Reddit is structured. And so they're funneling this the terror content onto Reddit.
They're working on Instagram. They're working on community notes. They're working on Quora, they're working on virtually every platform you can imagine. But I think in this case, they are very, very mindful, as I've said before, that they want this kind of material to stream downward, to Google, to ChatGPT or Gemini or whatever perplexity.
They know that this is going to end up there, because Reddit is so heavily indexed in these platforms that it's it's 1 to 1, like one post that gets enough likes or they call upvotes on Reddit. The material just gets pulled into the the LLM to the chatbot or onto Google. And when you ask it a question like who should I donate money to if I want to help children in Gaza?
You're getting this like terror linked charity that these guys had seeded upstream, because that's who they want you to give the money to, as one example.
00:32:01.910 — 00:32:06.750 · Speaker 3
Do we know who these Redditors are? Um, I know
00:32:08.230 — 00:32:31.370 · Speaker 3
you at least had some access to the discord servers that they're using. Um, do we know if there's if if this is an instance of, uh, foreign financing Iranian regime. All the bad actors who would, you know, the usual suspects or who would be funding this is. Is there some coordination there.
00:32:31.770 — 00:34:31.570 · Speaker 1
With regard to the the our Palestinian group on Reddit? We don't have evidence that there is. If there were if that were to turn up, it wouldn't really surprise me. Again, I kind of I, I draw a distinction between them and the Wikipedia group. We have found evidence of a group that's called tech for Palestine, which is at least supposedly an independent group that sprung up post October 7th or was founded by an Irish entrepreneur named Paul Biggar.
And that's where we found some open coordination off of Wikipedia and off of Reddit and other sites. Um, and on these discord servers. But in terms of like a state level operation, we don't have that at this point. I mean, this is kind of you get to you get to a point where getting to that level of, of access and information, more or less, I think, requires either an enormous journalistic task force that someone's just going to run for for months and months with a lot of people on it, a lot of resource, or you're doing some kind of government investigation where, um, you know, this is something we saw, I think it was in 2024 with a podcaster named Lauren Chen, who she wasn't actually named in the indictment, but, um, news reports identified her.
She was associated with a an operation run by Russia Today, which is the Kremlin, you know, Russian state TV, um, where she was being paid by them to help them disseminate content, whether or not she, she says she did not know that that's who it was, whether or not that's true. It's unclear. But the point is that it required a DOJ investigation to get to that point where you can actually subpoena people.
You get phone records, email records like this is the kind of thing that it takes to get to the point where you're able to identify foreign funding or some kind of coordination at the state level.
00:34:32.210 — 00:34:47.250 · Speaker 3
The AI companies, OpenAI, Google, perplexity, you know, they are the downstream people who are ingesting this. If the upstream companies can't can't police this, do they know about this?
00:34:47.330 — 00:37:55.910 · Speaker 1
Probably not. Um, I imagine that, you know, the people who are building the AI platforms, they are. So they're in an arms race with each other, and they're number one. And probably their only focus is to build better and better models that beat their competitors models. Like that's if they're not doing that, then they just lose and they die.
So mostly I imagine that it's not even on their radar. It's it's very kind of for them and side issue. Um, I think Google is a different case because Wikipedia exists as a function of Google. If Google didn't give it all that valuable SEO space, making it the number one search result on billions of searches every year, it would not exist as a website.
Wikipedia would cease to exist. All that traffic that flows through from Google gets monetized by Wikipedia in the form of donations that it solicits from users, and that's where they get most of their revenue from. They get donations and foundations that are giving, but they're also kind of stitched into Google in a weird way.
Um, you know, Google we think of Google as Google is today, which is this trillion dollar company. But it wasn't always that, um, you know, even in the 20 tens, it was a company that was still growing. There was still competition on search. Um, at least at the time. And what Google was able to provide through this partnership with Wikipedia was that when you googled something, the very first result that you got was guaranteed to be guaranteed to look to appear to be high quality and verified and neutral, most importantly.
So like if you googled a topic about like the history of the French Revolution and what you got back was some blog or let's say some some news site that had ads all over it or something that wanted you to pay for something. You just feel like this is just feels kind of like there's a motivation here to all the content and it's not it's not valuable to me.
But when you get a Wikipedia article that looks neutral, they're not asking you for money. They don't. They're not running ads on it. You're like, okay, this is actually a really good content. And Google got all of that value for free. They didn't have to pay for it. And at the time they probably couldn't have afforded to pay for it.
So Google and Wikipedia exist in a symbiosis, and it's a kind of a feedback loop. So Google is not entirely just downstream of Wikipedia. In a way, it's also upstream of Wikipedia. So I would hope that Google by this point, is starting to understand that Wikipedia is now becoming its own sort of liability to Google.
And the poisoning that we see on other platforms is happening to Google and by Google by virtue of this relationship, and that they start to say, is it really worth our time and our money and our risk and our brand and our reputation now that we are this world leading juggernaut, maybe there's a better way to solve this issue than Wikipedia.
00:37:55.950 — 00:38:09.660 · Speaker 3
You have uncovered in the past month. The what seems to be influence of the Iranian regime in Wikipedia stories about the current events going on in Iran.
00:38:10.780 — 00:38:18.220 · Speaker 3
How? How tight is that? Like how? How much of a state actor influences that explain what's going on there?
00:38:18.580 — 00:39:37.160 · Speaker 1
We have what we have with regard to the Iranian regime is a pattern of evidence that shows some kind of coordination in the direction of the regime. So the, you know, the edits, like a mention before removing human rights abuses committed by the regime, demonizing the MEK, which is the, you know, the this it's actually US designated terror group, but it is a threat to the regime and they hate it.
So you have sustained action on those pages. You have Iranian editors who are in control of extremely important pages related to the IRP, including page on Khamenei. Um, and you also have in most recently what we discovered was on Wikimedia Commons, which is where you upload all the different media files, videos, audio, pictures, that kind of thing.
A mass upload of at least 10 to 15,000 individual images and videos produced by Iranian state media. So we're talking about Khamenei IR, which is Ayatollah Khamenei personal official website. We have Tasnim, which is the IRGC s own owned state media company and major news agencies, which is another state owned media company, media organization.
00:39:38.240 — 00:40:45.540 · Speaker 1
These are their own produced images and videos. Some of them are watermarked. Some of them are labeled with with threats. Some of them show photos threatening the president of the United States or threatening Israel or other such such activities, and these are being uploaded by the thousands to Wikimedia Commons, and sort of becoming a distribution point of this type of material outside of their own little walled gardens in Iran.
So is that the Iranian government? To someone sitting in Tehran doing that, it could be I mean, it would like it would make sense. It could be some third party who's doing this for whatever other reason, I don't know, regardless of who's doing it or how the mechanics work, it's actually happening and it's happening with regard to editing on Wikipedia as well.
So we've been reporting on this in, um, neutral POV comm, that's POV Substack. And we'll continue to push that and push on it until we find more. And, you know, get our arms wrapped around the story. But it's definitely developing.
00:40:45.580 — 00:41:03.420 · Speaker 3
For somebody who is learning about this media literacy, who is trying to get a better handle on this sort of stuff. Is there a way to find out what is the good information, what's poisoned on? I mean, both on AI, but on on Reddit and Wikipedia directly.
00:41:03.580 — 00:42:14.980 · Speaker 1
There's not a good there's It's not like a turnkey solution to that. There's just trying to dig, to dig as much as you can. And whether or not that's, you know, in the power of most people in terms of busy lives and schedules and, you know, there's so many demands on us cognitively and on our time as well. So the idea that you're going to really dig into every piece is very difficult to do.
But I do think that there might be more. Some communities out there, WhatsApp groups where, you know, people circulate these kinds of things and they talk about it and they, um, they call this stuff out. Or, you know, groups like what you guys are doing at Honest Reporting, where it's getting called out, it's getting identified for what it is or what we're doing at POF.
Um, so it's about seeking out high quality sources like those, whether they're groups, whether they're organizations or companies or whatever, and, You know, trying to filter the news through those lenses. You won't be able to do that on every single case, but it'll sort of train you to how to identify a pattern that looks like it's something of that nature.
00:42:15.220 — 00:42:18.300 · Speaker 3
Um, if people wanted to learn more about your work, read more.
00:42:18.540 — 00:42:39.140 · Speaker 1
My my ex account, Ashley Rensburg, or our POV account on X is POV media. Um, or just come to neutral POV. That's the Substack where we publish all our work and um, all the latest stuff is there, including our the archive of other stories and other investigations. So we'll be happy to have you.
00:42:39.180 — 00:42:41.860 · Speaker 3
Awesome. Ashley, thank you so much for your time.
00:42:42.140 — 00:42:51.620 · Speaker 4
If you found this conversation useful, like this video. Subscribe to the channel and share it. For more investigations and analysis, visit Honest Reporting
00:42:52.700 — 00:42:59.380 · Speaker 4
and follow us on X and Instagram at Honest Reporting. This has been the honest state. Thanks for watching.