May 22, 2026
On May 22, The Capitol Forum hosted a conference call with investigative journalist Julia Angwin as part of our ongoing Tech Policy Briefing Series to discuss her recent New York Times opinion essay, “Meta Is Dying. It’s About Time,” which examines questions surrounding Meta’s long-term business model, platform strategy, and future trajectory. The full transcript, which has been modified slightly for accuracy, can be found below.
TEDDY DOWNEY: Hello, everybody, and welcome. I’m Teddy Downey, Executive Editor here at The Capitol Forum. Today I’m very pleased to be joined by Julia Angwin, an award-winning investigative journalist, contributing opinion writer for The New York Times, and founder of “Proof News” and “The Markup.” Julia recently wrote the essay, “Meta is Dying. It’s About Time”, examining growing questions surrounding Meta’s long-term business model, platform strategy, and future trajectory. Julia, thanks so much for doing this today.
JULIA ANGWIN: It’s great to be here. Thanks for having me.
TEDDY DOWNEY: And just a quick reminder, if you have questions, just enter them into the questions panel or the chat, and we’ll get to them later in the conversation. But to start off, Julia, can you walk us through your core argument in the essay? Why do you think Meta may be entering a period of structural decline?
JULIA ANGWIN: Well, I really started thinking about this when I was noticing Meta’s most recent earnings, where they announced their first decline in the number of total daily active users that they’ve ever reported. And it was a very small decline, but it reminded me of when I covered AOL back many moons ago at “The Wall Street Journal” in 2000. And they were in a similar point in their career where they were like really big. They seemed unstoppable. They had just bought Time Warner. But slowly, and under the hood, their core business was deteriorating.
And so, it started with a feeling of those vibes. And then I was like, okay, let me look a little deeper. And then when I looked into their financials, I was like, oh, it’s worse than I thought, right? Like, not only are they losing users, which means they own Instagram and WhatsApp, which are both doing really well. So, that probably means that Facebook is declining pretty quickly. And then when you look at what they’re doing with revenue, they’re cramming more ads per page because they don’t have as many page views, right?
So, they’ve actually increased what they call their revenue per user—which is basically the number of ads they show to a single user—by 27 percent in one quarter. So, that is a huge jump, which means they’re really in that moment where they’re really desperate to keep their earnings going up.
And then when you look at what the amount of debt they’re adding for AI, even Wall Street analysts are really getting nervous that this amount of spending is unsustainable. And that’s shocking given that they’re still hugely profitable and a huge company. So, if they’re piling up debt that Wall Street feels is unsustainable, that’s like a really big spending spree.
And when you add those together – you know, I have an MBA. I’ve covered business for a very long time. I’ve covered tech for a very long time. I’ve covered Facebook since the moment it was founded. I just was like this all adds up to something I’ve seen before, which is the way internet companies go out is they just become zombies. They don’t really die in a spectacular way. They just sort of like, they’re still—Yahoo still exists. It’s still profitable, weirdly. It’s like a shell of a company that’s just like zombies along, you know? And so, that’s what I felt like—I was like, I feel like I could call that for Meta.
TEDDY DOWNEY: And so, you make a bunch of points in the piece. As you mentioned, the users are declining. Facebook’s reaction is to stuff in more ads, charge advertisers more, effectively make the platform worse, that will exacerbate user declines. But beyond that, you also point out a lot of failed Meta initiatives.
JULIA ANGWIN: Oh, yeah.
TEDDY DOWNEY: That sort of implies Zuckerberg is effectively too incompetent to pivot successfully to another line of business, whether it’s AI or whatever else might be. Can you talk a little bit about the sort of the failure to pivot or augment the business or?
JULIA ANGWIN: Yeah, any one of Zuckerberg’s failures would have doomed any other CEO. But because he has voting control of the company and can’t be voted out by the shareholders—because of his sort of super majority voting stock—that he has survived what would doom anyone else, right?
So, like, just one of his disastrous investments we could talk about is the metaverse, which was this big, huge bet on virtual reality, which he poured more than $80 billion into. And the idea was we’re all going to wear these weird headsets and like be avatars and go walk around in a virtual world where no one had legs. And it was just a complete disaster. And from basically 2021 to 26, he just plowed absolutely gob smacking amounts of money into this thing when it was very clear from the beginning, like no one was using it. It was an empty shell all along. I still remember the great story that Kashmir Hill wrote in the New York Times where she went and hung out in the metaverse and she was like, there’s no one here.
And so, that would have been enough, I think, for most CEOs. But because he’s really not accountable to shareholders, he’s now done another insane amount of spending on AI. So, he spent almost $100 billion, I think, and that’s probably like a low estimate, on jumping into the AI race. And he put all of his money and his sort of moral persuasion into this idea that he was going to build this open source AI model, Llama. And that unlike ChatGPT and Claude that are closed source, he was going to offer this like democratic version of an AI model that anyone could run on their own desktop.
So, the idea is you really would control the AI yourself, right? Instead of just like what we do when we’re using ChatGPT, we’re really connecting to open AI servers and running it on theirs and they can see all of our traffic. With Llama, you were going to be able to control it. The only problem is Llama sucked and it didn’t compete. And also, it was really heavy to install. Like it actually wasn’t usable for most people to install on their machine. It wasn’t something anyone wanted to install.
So, after losing $100 billion on that, he fired that whole team and poured another $14 billion into hiring a team who’s never built a model before and brought them in and now has told Wall Street is going to spend minimum $115 billion just in the next year on this new closed source model that’s going to be amazing.
And look, he did manage to launch that model earlier this year and it like performed okay. It didn’t perform great, but it was like it was better than Llama. But like that’s about it. Like he’s still way behind on this AI race and he’s borrowing money to do it in a way that Wall Street is really worried about. So, it just feels like a real—just one disaster after another.
TEDDY DOWNEY: You touch on the debt spending and AI. But one of the things I think that’s dogged Facebook recently is they don’t have a cloud business, which is one way that Microsoft, Google, and Amazon are betting from the big, huge boom in AI spending. So, they have a lot of CapEx, but they’re taking in a lot of money.
JULIA ANGWIN: Yeah, exactly.
TEDDY DOWNEY: So, if you don’t have a cloud business, that means you just potentially get stuck with that CapEx.
JULIA ANGWIN: Yeah, exactly. You can’t do what the other companies are doing, which is like—I hate to use this word, but it’s a little bit of a round trip, right? Where they spend and then get back on the cloud spending.
TEDDY DOWNEY: But do you think that is another layer of this problem for them?
JULIA ANGWIN: I mean, structurally, that’s a problem, right? I mean, but it speaks to a larger problem, right? Which is that Meta is a consumer business and the AI business right now is a B2B business, right? It’s a businesstobusiness enterprise business. And Meta is not an enterprise software company. Never has been, never will be.
And one thing I know from having covered software my whole career is that it’s really hard to change. If you’re an enterprise software, it’s hard to make yourself into a consumer software company. And if you’re a consumer software company, it’s really hard to make yourself into the other.
The only one that I really know has done both medium well is maybe Google. I would argue Microsoft has not done both well, even though I think they would argue they have. But ultimately, Microsoft’s core is enterprise. Google’s core is consumer. And Google amongst those is the only one that’s really managed to have like a significant cloud business, which is an enterprise business.
And so, when you talk about a company like Meta trying to get into cloud or AI, which is mostly an enterprise business, it doesn’t fit with their model. It doesn’t fit with what they know how to do. And so, I think it’s sort of not a natural fit. It’s not something that I would look at and say, I think they’re going to have a good chance of winning here. Because these are entirely different businesses selling to business or selling to consumers.
TEDDY DOWNEY: One way that Facebook has survived over the years—you mentioned Instagram and WhatsApp are doing well, obviously, also owned by Meta—is they just buy their competitors. Even if they haven’t innovated at all, really, they’ve bought two very successful close competitors. Can’t they just keep buying their closest competitor and just force you into this world?
Because I think that’s part of how they’ve been able to keep going as long as they have. And I guess that’s a question also. Are you talking very specifically about Facebook, the app, or the whole company, when you mean like Facebook is dying?
JULIA ANGWIN: Well, Facebook is dying. But I think it’s going to bring Meta down with it, is really what I’m saying. Because Meta’s all three products, really Facebook, Instagram, WhatsApp, but Facebook is still the vast majority of their ad revenues. And ad revenues is all of their revenues.
So, ultimately the core of their business is still Facebook, right? WhatsApp is really not particularly monetized. It’s not that significant. Instagram is. But it is still smaller amounts of revenue. And it’s growing quickly much, but it still doesn’t compete with Facebook’s ad revenue. And so, the reality is they have kind of this eroding core and these two smaller products that are. One of them does have a revenue trajectory that might be more meaningful, but WhatsApp doesn’t as much. And they have bought their way into success, right? Instagram and WhatsApp were their ways of buying competitors. But the thing is, those are just the ones that worked out. They’ve bought all sorts of other companies that didn’t work out, right? They bought a headset company for their metaverse thing. They’ve bought various AI companies. They just bought another one to do this latest AI thing.
And what people don’t remember about Facebook is how many companies they’ve bought and then have died basically in their hands, right? There’s this thing people talk about, which is like the catch and kill model of these companies, where they buy these companies, almost really not even to make them succeed, just to take them off the field, right? And so, they’ve done a lot of that, but we don’t—actually, I don’t have a list of all their acquisitions. But if I did, I would tell you probably the vast majority have been unsuccessful.
TEDDY DOWNEY: And you mentioned TikTok in the piece, you mentioned it’s not quite as bad as Facebook in terms of pushing divisive content. I was curious, is there actually a lot of evidence of that Facebook is pushing you into more radicalizing content than TikTok? I’ll admit I haven’t really used either. So, I don’t really know.
But certainly, being in Washington, we hear a lot of complaints about TikTok, Facebook, YouTube, YouTube Shorts, whatever. They are all kind of are pushing towards extreme content. But is there a lot of evidence around TikTok not being as bad or the algorithm working differently?
And then I kind of want to talk a little bit more about, is it really a legit alternative? Or what do you see as the legit alternatives to Facebook?
JULIA ANGWIN: Well, I think the one thing we have a lot of evidence about is we have a lot of evidence, at this point, about how Facebook tunes its algorithms towards what we call engagement bait, right? And from the Facebook files that were released by Francis Haugen, we know other whistleblowers. Also, there’s been rounds and rounds of sort of evidence that have come onto the scene about how they promote content that basically will keep you on the site longer.
And they know that outrage is one way to keep you there. And so, there’s absolute evidence that these algorithms are tuned towards outrage and things you can’t look away from. It’s also true, by the way, that people like to look at puppies. And so, it’s not like the engagement algorithm doesn’t only foster outrage, but it really does have a huge outrage component to it, right?
So, that’s what we know about Facebook. TikTok algorithm is much less studied, much less knowable. And so, we don’t know as much about it. It’s a black box. It is also very different than Facebook’s algorithms and Instagram’s algorithms. Because TikTok is all about showing you new content. The sort of premise of TikTok is discovery.
So, it’s like, I’m just going to show you something you would never have thought of before. And it’s going to be a new dance or a new song or a new something. And that’s why people love it. They don’t have to make that decision of like who to follow or whether I’m following the right people, but they just are served up cool things.
And so, one thing about TikTok is their algorithm is sort of optimized more for this idea of new stuff, which means you’re as likely maybe to keep getting the same outrage and going down the rabbit hole. Now, I can’t prove that because it is difficult to analyze the TikTok algorithm. And there’s only been a little bit of work on it. Also, really hard because it just got bought and it’s changing hands. So, we don’t know what the algorithm is going to look like in a minute.
But if you just look at the lists of what were the most popular things on TikTok and the most popular things on Facebook, you will just see a huge difference. Like, I don’t know that there’s been any, you know, ever since Facebook bought and killed CrowdTangle, which was the only sort of analytics tool we had for what was trending on Facebook, we have very little insight.
But back in those days, when we had some insight into what was trending on Facebook, it was all outrage content. It was The Wire and Ben Shapiro and just sort of people who were really interested in sort of shouting about bad things happening. And whenever you see a list of hot things on TikTok, it’s like a dance craze and a cute grandma making cookies or whatever. So, I’m just somewhat anecdotally saying that so far it doesn’t seem that it’s tuned exactly in that same way.
TEDDY DOWNEY: Another thing that I’ve seen is Facebook’s conduct seems pretty egregious in terms of how they are training their AI, in terms of how they’re getting the content. Do you think they have any significant copyright litigation risk? Or have you looked at that at all in terms of another risk to Meta?
JULIA ANGWIN: Well, I think they have the same risk that everybody has, which is they’re all using copyrighted content. I think Anthropic just made a huge settlement with the authors. Llama, as open source, actually had—it was a little bit more transparent about some of the data it was being trained on. And it was the same as everyone else’s, which was basically a lot of content that was obtained without consent.
And so, the copyright lawyers will have to go into court and make the legal case about whether obtaining it without consent was illegal. But it certainly happened without anyone knowing that it was going to be used for this reason. So, I think their legal risk is equivalent to the others, because it was trained the same way. Now, I don’t know this new closed model, whether it’s a different thing. But ultimately, all of these bots are training off the same data sets.
TEDDY DOWNEY: Yeah, yeah. I mean, we’re tracking that pretty closely. It has a big effect on the media industry generally, if they can call that fair use or not.
JULIA ANGWIN: Yeah, exactly.
TEDDY DOWNEY: So, one thing I want to touch on—and then I want to talk about your book. But generally, social media and AI are becoming increasingly unpopular. You don’t really discuss this too much in the piece, other than to mention Facebook faces billions in liability on personal litigation in court. I think that kind of mixes in with that. We have mentions recently of AI being booed at college graduation speeches, social media being banned across the world for people under 16. I mean, we could point to a lot of different things. But there seems to be a cultural backlash. Do you think that that is a problem for Facebook and Meta in terms of people using the apps? And then also in terms of continuing their winning streak in Congress and in the courts?
Now, obviously, you mentioned they’ve lost recently in court. But generally, they’ve done well in court, particularly on their antitrust case. What do you see as the most important impacts on social media and AI from them just becoming broadly unpopular?
JULIA ANGWIN: The funny thing is that I think they’ve been broadly unpopular for a while. It’s been a pretty bipartisan feeling about Big Tech having too much power and needing to be reined in. But because Congress has failed to do anything with that bipartisan energy it’s been left to the courts. Because the citizens don’t have that much ability to get off of these platforms.
The reality is that, like, you can hate social media as much as you want. But like, if you want a job, you have to be on LinkedIn. And if you want to talk to your grandma, you probably have to be on Facebook. And they make it really hard for you to extract from these networks and bring your network with you to a new platform, right?
So, we as users are locked in, in a lot of ways that means that basically, even if we don’t like it, we don’t get to vote with our feet as much as we would like to. That said, I think the user number decline at Facebook indicates that people are leaving, even though it probably costs them.
And I think that the court rulings, certainly these personal tort cases are—I think, the fact they had this bellwether win against Meta could easily usher in, right, this huge tobacco-sized settlement. I think everyone sort of expects that will eventually happen as a roll-up of all of the individual cases. And that could be the—I think the tobacco one was in the hundreds of billions. And we could see something like that for Meta maybe as well.
And I think sometimes it’s worth just thinking about the courts and Congress as like different ways that we can express our views as citizens. Because we can’t vote on this, right? Like there’s not like a candidate that I know of who’s running on the anti-social media platform. Although, I feel like maybe now that’s starting to happen a little bit. But it’s still not like the same kind of wedge issue that you can like win or lose on. And so, our ability to express our views is limited. Congress hasn’t been really acting on behalf of people’s desire for more reining in of these technologies, but the courts are. I think partly because they’re just frustrated by the impunity these companies have. There’s so much evidence out there, right? Like at this point, we have 10, 15 years of investigative journalism showing not only that the harm was created, but that they knew it was created and decided not to do anything about it. And that I think is the thing that the courts are acting on, which is like the negligence question, right? You knew this was going on, but you decided just to ignore it.
And so, I think that the sentiment we’re seeing bubbling up is it’s playing out in the courts because that’s the only place we have an avenue for that. And I would just say that is a symptom of the fact that our democracy isn’t working that well, right? Because we should have more ways for something that is so popular. And when you ask people, what do they hate most in the world, it’s like AI is up there with the Iran war. It’s like, okay. These are the things we really hate. And somehow it doesn’t filter out into our legislative environment. And so, that’s a failure of democracy.
TEDDY DOWNEY: I want to get to that because I think that has something to do with your book.
JULIA ANGWIN: I know, watch me try to tee up my book. But it’s also true, I just want to say.
TEDDY DOWNEY: I’ve got one last question or maybe two last questions, then we’ll get to the book. You make the point that as Facebook declines, it can have really significant societal effects, even worse potentially than when Facebook has been doing well, which, as you mentioned, that was pretty bad in terms of radicalizing people, sparking genocide, swinging elections. How could things actually get worse?
JULIA ANGWIN: I know. And I was really sad to write those paragraphs. Because I really wanted to be able to just dance on the grave, right? Because I think we all feel a little bit of like, goddammit, this company has been really just acting with impunity for so long.
But the reality is that when companies are in decline, that can be their most dangerous moment, right? Like I think I mentioned in the piece, Yahoo, they were declining, they had made a whole bunch of acquisitions in the hopes of competing with Google, but Google just really trounced them. And they sort of under-invested in cybersecurity and had the largest data breach of all time. It was also state actors. I believe it was Russians. And so, they basically lost every user’s data to adversarial state actors, right? And that is really bad.
And unfortunately, Meta has been cutting back on safety and content moderation, all the things that supposedly keep their platform safe. And so, I do worry that they are going to start letting child sexual abuse material circulate. Or they’re going to be turning even more of a blind eye to harassment and scams. They’re already—the FTC report, I think, that recently came out, said that they’re the number one spot for scams. And like more than half of all scams reports to the FTC happen on Meta properties.
So, it’s already kind of a place that’s like not well policed. And then as they struggle to keep meeting their profits, I worry that they’re going to cut back even more on any of the safety measures they have. And so, that means it’s just more of a dangerous place to be, right? That you’re going to be more likely to be scammed there, that you’re more likely to be exposed to horrifying material. And also, that more false narratives that are divisive and perhaps leading to terrible political violence could also be circulating there. So, I think there’s a lot of bad things that unfortunately can happen when a company is in dire straits.
TEDDY DOWNEY: Again, for our listeners, if you have questions, please put them in the questions panel or put them in the chat. In terms of solutions, what do you think should be the top priorities for people who actually want to solve the problems here? State AGs, state legislatures, foreign governments, members of Congress, especially those who are potentially bringing new leadership in November. To your mind, what are the best solutions to dealing with the problems that Facebook is presenting?
JULIA ANGWIN: I mean, I will say this. Look, I don’t know that I have all the answers. But I can tell you a few things, which is that I think the very bare minimum that this country needs is a privacy law. Every other nation basically has a comprehensive privacy law that sets minimum standards for the treatment of personal data.
And as I mentioned, I worry that Meta has a lot of data on us. And if they don’t protect it, we are all going to be very much exposed and vulnerable. And so, I think at the very least, we have to have a minimum comprehensive federal privacy law.
I think we also need—in all cases, I think basically the answers are transparency and due process. So, I think we need to know a lot more. It’s sort of like what we have with the EPA. We need—companies need to say how much they’re polluting. So, Meta needs to publish all its harms and do its own audits. But then the EPA also doesn’t just trust companies to say how much they’re polluting. It also measures the water and air itself. So, we need outside auditing and outside accountability.
And then we need due process. So, if you’re harmed, you need to have some recourse. Right now, there’s no recourse. Anything happens on Meta—your whole life savings are scammed from you, whatever—there’s nothing you can do. They won’t even answer the call. There’s no phone to call, right? And so, I mean, I don’t know that I have the perfect solution for this, for Meta. But I would say those three things are like the things you need for a functioning like relationship between a corporation and people.
And then I would say, specifically on social media, I am a fan of the federated social media model. Like the idea that Blue Sky and Mastodon have of federation, where you don’t have to be locked into a platform, that you can bring your friend network with you to different social networks, and that that is portable, I think is a really good solution to the fact that I think we still do want some of the functionality of social media. We want to be able to broadcast.
Basically, social media makes everyone a media property, right? Everyone has the benefits and the downsides of that. But the benefits are there are a lot of people who did not have a voice before. The media was not amplifying their voices. And social media gave them that. And I don’t want that to be gone, right? I’m not one of these like kill social media people. I think we need to just evolve it for the modern era. And the federated decentralized model is the best one I’ve seen so far that looks like it’s the way forward for the future.
So, I am a big fan of Blue Sky. I really want that whole model to succeed. And I think that there should be a million more competitors in that space who are all connected to each other through this protocol, and allowing you to choose which one you want to be on. And so, I hope that that succeeds so we can have the benefits of the social media era without the downsides.
TEDDY DOWNEY: And anything on the antitrust front? Do you think these companies are just too big and powerful at this point, need to be broken up? Or are you sort of more like if they just had these rules, they wouldn’t be able to do all this bad conduct?
JULIA ANGWIN: I mean, I think they also need to be broken up. So, in addition—sorry, I forgot about that just because it’s been so politically taken off the table. But I wrote a piece for the Times after the Google antitrust trial. It was such a disappointment, right? To have the judge rule they were an illegal monopoly and then say, but we don’t need really meaningful remedies because AI will provide enough competition, was so disappointing. Especially given that who do they think AI is? It’s the same company. It’s Google, right? Like Google is already been using its power in search to dominate AI.
And so, like already the thing that the judge had said was going to bring competition has been quashing competition, right? There’s a reason Anthropic is mostly an enterprise play in AI because Google’s crowded out everyone in consumer space. And there are some data that show that Google is Gemini. Because it’s in Search, it’s already more used than ChatGPT.
And so, I think we have seen a real failure of imagination in antitrust—and even in cases where the government had gone to the great lengths to prove that Google was an illegal monopoly and won that historic decision. I’m still really crushed by that, to be honest. And so, I think I would love to see these companies be broken up and in meaningful ways, right? Like not just like, yes, we should spin off Instagram and WhatsApp. But I actually think that these days, the monopoly over data is the piece that has to be broken up. So, you also can’t maybe have all the ad data and all the audience data, right? Like there’s too many data sets. I think we need to start thinking about antitrust in terms of which databases need to be broken up, not just which consumer facing products need to be broken up.
TEDDY DOWNEY: Yeah, I can’t even keep track of all the different types of data that Google has, right? You’ve got Waymo. You’ve got Gmail. You’ve got Search. You’ve got ad tech. You’ve got YouTube. I mean, you just spend all day thinking about how incredible, how powerful all their data is.
JULIA ANGWIN: Yeah. And it really comes to a point that I think I told you before that I want to make, which is I think at this point, we need to start thinking about Big Tech as big media. Because everything you just said about Google is access to information. If Google wants to change the name of a country, for instance, or remove a river from the map, for all intents and purposes, that country or that river no longer exists. Do what I’m saying? It obviously exists if you drive there. But their mediation of our information landscape is their political power. It’s their financial power.
And so, one of the things I’ve been thinking about a lot these days is that we really need to rename Big Tech as big media. Not they don’t have tech. They do have tech, obviously, undergirding their work. But the real power that we need to focus on, that is affecting all of us, is their mediation of our information environment. And that is the role that media used to play, right? Media now—the media that I grew up in—The Wall Street Journal, ProPublica, et cetera—that is intermediated by tech, right? They can’t even get their stories out if the big platforms don’t let them.
And so, I think we have to remind ourselves of where these companies have their political power, how they wield it, and then think about also the regulatory remedies with that frame in mind, right? Because what we really want is to make sure they’re not ginning up the playing field for certain types of information to be suppressed or some others to be amplified, right? That’s extremely dangerous to a society.
And if you don’t mind me getting ahead of my book a little bit, that’s what authoritarians want to do. Authoritarians want to control the media environment. That is their main game. Ninety percent of their effort goes into controlling media.
And so, it’s not a surprise that the Big Tech companies, the CEOs, were all at inauguration this year. That is what an aspiring authoritarian, like our current leader, does. They need to control the media. They need to bring them in close and make them feel that there’s no other option than to cooperate. And that’s very dangerous for our society.
TEDDY DOWNEY: It’s funny that you say that. Just dealing with the Trump administration press people, they’re constantly calling and yelling and complaining. And, I mean, there’s some level of like press people that do that, but there’s a certain twist to it that feels like they’re policing the media in a much more heavyhanded way. It’s funny that you say that that’s kind of a staple of authoritarianism. I know you’re writing a book about authoritarianism, how to fight authoritarianism. Can you tell us a little bit about what you’re doing there?
JULIA ANGWIN: Yeah. Thank you for asking. Now that I’ve forced you into this. But I do believe it’s really relevant. So, I’ve written this book called “On Courage: How to Be a Dissident in an Age of Fear.” It’s coming out June 30th. And basically, I woke up the day after the election and was like you know what? I’ve spent most of my career writing about corporate power and unaccountable corporate power, mostly in the tech industry. And I was like you know what? I really need to start focusing on authoritarian power.
And so, I teamed up with my coauthor, Ami Fields-Meyer, and we basically interviewed more than a hundred dissidents from around the world, inside and outside of the U.S., basically about what does it take to fight authoritarianism? What are the stories? What does it mean? How do we do it?
And the book became really inspirational to me because I think I felt like a lot of people are sort of hopeless. But once I reported out all these stories, I realized people have been fighting for their rights in much worse conditions than we have right here and have won. It’s not easy. It’s not simple. But it’s possible.
And so, it made me remember that we have the tools to do it, and the tools are actually each other. The tools are humans connecting with other humans around shared values and demanding that their institutions and governments respect those values.
And the funny thing about a lot of anti-authoritarian movements is they’re really simple. They’re like the Soviet dissidents, the only thing they were asking for out of the Soviet Union was like we have a constitution. Just follow it. I feel like that’s a little playbook for us. We have a constitution. How about we follow it?
And so, I think we don’t have to look hard. Everyone’s like, what are we going to align around? I’m like, I don’t know. We already have this totemic document. Let’s just go with what we have. And I think No Kings is a very good framing for that because we were founded on the premise of No Kings. The Constitution was written as a rebuke to kings. And so, I think we’re on the path already where people are starting to unite around this idea of we do have a shared value in this country and let’s insist on it.
But I think that it just really was a really wonderful exercise for me in learning all these stories. And I hope that readers will find it inspiring. It’s funny to have been an investigative journalist who has basically spent their whole career writing depressing news. Like here’s a terrible thing that’s happening that you didn’t know about. Here’s another terrible thing that’s happening you didn’t know about. This might be the only, the first hopeful thing that I’ve written.
Although, I will say a lot of people felt the Meta piece was really hopeful. They were like, oh, could it die sooner? So, maybe this is also part of my hopeful agenda. But anyway, so “On Courage” is really all about how to fight back and how you should believe that it is possible.
TEDDY DOWNEY: What kind of a role does tech, Big Tech, big media, play in allowing for authoritarianism? Because most people in this country think—I think wrongly—that this is a functioning democracy. Princeton did a study saying, no, this is oligarchy or aristocracy or whatever they called it. I think there’s so many examples of the government not being accountable to the people. And, I mean, we could spend hours talking about this. But I guess (a) do you think we have authoritarianism here? And if yes—which I’m guessing is yes—what kind of role has Big Tech played in allowing for that?
JULIA ANGWIN: Yeah, what we have, according to the experts, is democratic backsliding, which is sort of falling in towards authoritarianism. But we’re not all the way there yet. So, that’s good news. Which means that there’s time to turn it around. And, of course, there’s still time to turn it around, even if you go slide all the way there. If you look at Hungary, right? They just did what’s called the democratic U-turn that the people who track this thing talk about.
So, there’s always hope to go back because authoritarianism is also very brittle, right? If you’re not actually responsive to the people and you’re just trying to oppress them into doing what you want, ultimately, that’s a very fragile situation because there’s more of them than you, right? Like just the real facts, right? It’s a numbers game and you can only kill so many of them, right? And so, that’s sort of the game of fighting authoritarianism is you just have to get it to be clear to the government that there’s so many of you and they probably need to give up on trying to do all the bad things they’re trying to do.
So, the way that tech fits into this, though, is disturbing, right? Because one of the things that we are facing right now is a lot of the historical lessons that we lean on—like the Soviet dissidents or the Montgomery Bus boycott, like really effective moments of non-violent civil resistance that have changed the world—they were not happening in the sort of tech-fueled surveillance states that we have built, right? And so, that is a very disturbing thing. We have two chapters in the book about basically this idea of like it’s very hard. Because your devices are tracking you. We’re in this surveillance world. And what are the effective strategies for dissidents?
And the answers were unsatisfying, right? You can do a lot. You can use a burner phone. You can use signal. You can do you can do a lot of countermeasures. But in the end, like the government has drones with facial recognition and they just have so much.
And so, ultimately sort of the answer we had was that (a) you can try—you have to be smart about your use of tech because you have to understand that it’s a tool they’re using to surveil you. But that also the real defense is other people, right? We saw this in Minneapolis, right? It wasn’t tech that had protected people. It was people who protected people. They put their bodies in front of the people that the government was trying to snatch off the streets. And, I mean, it’s dark to say that, and I don’t want it to get to that point. But that is sort of ultimately the best defense is other people. And the more people we can—that work together to defend each other from state violence is actually the way you win against authoritarianism.
TEDDY DOWNEY: And you have a new center at Harvard—we talked about this a little bit—to promote independent media. What are you doing there? And what do you think is the role of independent media in all of this?
JULIA ANGWIN: So, one of the things I realized, while researching all of these authoritarian regimes past and present, was that authoritarians are so obsessed with capturing the media that controlling the narrative is like 90 percent of the game. I think I went into this reporting thinking it was maybe 50 percent of the game, but it’s like 90 percent of the game. And the other 10 percent is like state violence and gulags and repression.
And so, once you realize that—and also you realize that basically corporate media almost always caves, right? It looks different in different countries. In Hungary, Orban just defunded independent media and put all of the state—put state advertising money into the big broadcast legacy media, and then made them into arms of his state propaganda. In Russia, Putin just sort of basically killed a bunch of journalists and scared everybody straight.
So, the capture can look different in different countries. In the U.S., we’ve been having this crazy consolidation of power where we have one family, the Ellisons, who’ve been buying up all of the leading media properties. So, that is a different way that it can look. And this is a family that has made it very clear that they’re on Trump’s side.
So, in every country that we’ve looked at, independent media has been basically the only lifeline for the antiauthoritarians, right? Like whether it was the newsletter, “Samizdat”, that the Soviet dissidents were distributing, or the few news outlets that kept going in Hungary, there’s always sort of an independent media that is willing to still be fearless and adversarial. But usually, they’re smaller because the big ones get captured.
And so, this center is really my attempt to focus on independent media and support them. I think we don’t have – a lot of the way that academia has looked at what I call independent media space is your outlet, obviously, but like YouTube, TikTok, Substack, this is where people are going, right? They’re not going to the evening news anymore. They’re all on these other platforms. And a lot of what appears there might not even look or self-identify as journalism, but it is still where the narratives and the public discourse are circulating.
And so, I want to take that area seriously. A lot of work has been focused on tracking disinformation in that space. And that is important. It’s important to know what lies are being spread where, but it’s sort of impossible, right? That there’s an infinite number of lies, especially now that AI has powered it.
And so, what I’d like to focus on a little bit more is the quality. Where’s the quality information? How can we support those quality networks? I’m working with a bunch of creators on building some shared ethical guidelines and standards they can all say to their viewers, like we hold up these standards.
I’m working on building ways to track narratives that are traveling across these spaces because no one is really tracking that. So, I’ll hopefully be coming out with some reports fairly soon about like what kind of narratives are trending on YouTube.
And so, I want to take this space seriously and support its evolution and treat it like the real news source that it has become.
TEDDY DOWNEY: And do you distinguish between—it sounds like you’re talking about both the content creation and the distribution channel together, right? That media ecosystem, like what’s going through the media. Because YouTube is more like a distribution channel and people are making the content, right? Versus Warner Brothers where they’re making the movies. But yeah, I guess they have a distribution channel too and Paramount Plus.
JULIA ANGWIN: Yeah,
TEDDY DOWNEY: But obviously you have movie theaters and they’re selling the content to other people as well.
JULIA ANGWIN: Yeah. I mean, one thing that’s important to recognize about all media, but particularly independent media, is at this point, all of content creation is platform dependent, right? It’s dependent on these companies that we used to call Big Tech that I’m now calling big media. They are the gatekeepers. They choose what we all see in our feeds, right?
And so, one reason I want to support the creators is I want, by measuring what’s happening on these platforms, through their eyes, essentially through the lens of what content is circulating, I hope that we can actually answer questions like is YouTube suppressing certain types of content? Is TikTok suppressing certain types of content? These are the kinds of questions we’re going to need to know, right? Because whether it’s happening now or not, it will be something that an authoritarian regime will try at some point, right?
And so, we need to know a baseline of what’s circulating now so we can understand if it changes in the future. So, one aspect of this is trying to defend and understand what content is being created. And then understand if the platforms aren’t letting it through, that’s something we need to know about.
TEDDY DOWNEY: At some level, don’t you know that because they’re prioritizing rage bait or whatever, outrage content, necessarily they’re going to downplay your Capitol Forum 10-page, in-depth, article about some kind of anti-competitive conduct. You know what I’m saying? Or what have you, if we were to do a —
JULIA ANGWIN: I mean, maybe. You know what’s weird though? If you look at YouTube in particular, and I would say YouTube is the one that I’ve focused on the most because YouTube is bigger than TV ever was. YouTube is like what broadcast TV was in its heyday squared, right? YouTube is the biggest media platform in the world.
And if you look at what’s popular on YouTube, yeah, there’s some rage baited stuff. But honestly, it’s actually really heartening to see that really deep explanatory work about why is this thing happening in the world is actually quite popular. I think that we’re getting a little bit of a sense through YouTube of what the audience wants, and it’s actually not exactly what we thought.
They do like deep dives. You know what they really like? Two-hour long, three-hour long videos, right? They like long videos. They like experts speaking to them directly instead of being intermediated by a journalist. They like really deep analysis. They like maps.
So, there’s all sorts of things you can learn, and I think that is really heartening. Of course, there’s a lot of nonsense that people like too, but that was always true. People have always liked nonsense. It came in different forms, you know?
TEDDY DOWNEY: Yeah, yeah, yeah. I guess my interaction with YouTube is with my kids or seeing other people’s kids, and they eventually end up with creepy videos or short, super short videos that are sort of killing their attention span. Look, I obviously haven’t used it enough to be an expert on it by any stretch of the imagination. But it’s interesting to hear what you’re saying because I have heard that people, if you are really intentional, you can learn a lot. Like let’s say you have a hobby or what have you and you want to follow someone who’s a real expert, but that you’re wading through a lot of other stuff to get to that.
JULIA ANGWIN: Yeah. Well, this is sort of where I think that’s where I want to have a little bit more algorithmic accountability, right? Which is why I’m doing my sort of analysis of these platforms because they could choose to push whatever they want.
And the fact that these high quality explanatory videos get as much views as they do—maybe without even the boosting of the algorithm—shows how much better they would do if the algorithm was boosting them. But I think that YouTube is big enough that you can pretty much find anything on it.
And so, I think one of the things I want to sort of understand and do research on is who is seeing what, right? Like, the high quality stuff that I’m seeing, like that has 20 million views, right? These are huge, huge platforms. But is it really only just being shown to certain segments?
And that part we don’t know. The audience behavior data is something these platforms have a monopoly on. They do not share it. They sell access to it to advertisers in a way that the advertisers themselves don’t know, right? Like the advertiser comes and says, I want to buy a bunch of like men who have kids and whatever of this age. And then the platform will say, okay, we showed it to those people. There’s no independent assessment of whether that actually happened, right?
So, one of the things that I really am focused on is they have an absolute monopoly on audience behavior data, who’s watching what? Something that we used to know. We used to have a Nielsen, right? Told us soccer moms are watching this thing and whatever. We don’t have that anymore. We don’t have any independent assessment of who’s watching what. And that’s actually a real detriment to us as a society.
And we need to know that information because that information is what allows for good information to circulate. How is a person going to send out information about an important thing that the community needs to know if they don’t know where to find anyone, right? So, those are kinds of things I’m trying to do research to reveal.
TEDDY DOWNEY: Well, Julia, amazing essay. I’m super excited about your book. What’s the name of your book again? And when does it come out?
JULIA ANGWIN: Thank you. It’s called “On Courage: How to Be a Dissident in an Age of Fear.” And it comes out June 30th, but you can pre-order right now on your preferred platform.
TEDDY DOWNEY: Excited to pre-order it, read it in real life. And I’m excited to see these studies out of your new center at Harvard.
JULIA ANGWIN: Thank you.
TEDDY DOWNEY: So, thank you so much for doing this today.
JULIA ANGWIN: Thanks for having me.
TEDDY DOWNEY: And thanks everyone for joining us today. This concludes the call. Bye-bye.
