The Vergecast - How pandemic disinformation is affecting social networks

Episode Date: May 12, 2020

Alex Stamos, former chief security officer of Facebook and director of the Stanford Internet Observatory talks to Verge editor-in-chief Nilay Patel and editor Casey Newton about his role consulting on security for Zoom, disinformation around the pandemic on social media, and the threat of foreign interference in the 2020 election. Learn more about your ad choices. Visit podcastchoices.com/adchoices

Support for the show comes from Retool. Too many companies run critical operations on duct taped spreadsheets, Slack workflows, and whatever else they could cobble together. Not because they want to, but because building internal tools means weeks of waiting on someone else's backlog. That's where Retool comes in. Build custom internal tools just by describing what you need. Prompt something like,

Build me a revenue dashboard on our Salesforce data. And Retool actually builds it on your company's data, in your cloud with enterprise security built in. Go to retool.com slash Verchcast. We all need to retool how we build software. What's up, y'all. I'm Skylar Diggins, seven-time WMBA All-Star, Olympic gold medalist, and mom. And I'm Cassidy Hubbard, host and reporter for nearly 20 years, covering the biggest names and stories in sports and mom. And this is Am Mom, a community for athletes, game changers, and moms of all kinds.

dropping May 14th. Tap in with us. Hey everybody, it's Seattle from the Vergecast. On this week's interview episode, Casey Newton and I talk to Alex Stamos. He's the director of the Stanford Internet Observatory. He was also previously the head of security at Facebook. Alex is all over the place now. He just took a job consulting on security for Zoom.

We talk about that. He's really interested in how disinformation around the pandemic is playing out across social networks. We talk about that at length. And then we get into the election. Obviously, in 2016, there was a lot of, of foreign interference on social networks with the election. Alex and Casey and I talk about that in detail. It was a super interesting conversation.

Alex is not shy about this stuff. And he has just an incredible grasp on all of the different vectors that you have to think about when running a social network at scale. Check it out. Casey Newton, Alex Stamos on the Vergecast. Alex Tamos, you're the director of the Stanford Internet Observatory. Welcome to the Vergecast. Hey, thanks for having me.

And I'm joined by my friend Casey Newton. Hey, Casey. Hi, Eli. So, Alex, we've had you on a couple times before. We've often talked about social platforms, democracy, all the stuff. Casey's interested. You were the head of security at Facebook for a while before that Yahoo. Now you're at Sanford. You also just recently, and I think this is super interesting, took a job advising Zoom on their security issues. I also want to talk to you about pandemic misinformation. There's also an election coming up, which seems like it has some security implications. So a lot to talk about. but I want to start with Zoom. Zoom has been just in the press a lot. It's obviously how explosive growth. It's been criticized for security practices.

How did you end up advising them on security? So this is the first time that Twitter has been good for my career, actually, in that I was criticizing Zoom on some of their issues and talked about how this was a need, that they were demonstrating a need to have like a real security turnaround. So I had some, you know, thoughts I threw it on Twitter. and then the CEO, Eric Ewan, went and found my phone number from a friend and gave me a call and said, asked me what my advice is. We talked for like an hour and a half.

And he ended up asking me to advise him. So what does that look like? Are you in a counselor or a board of advisors? It's just you get free reign. Are you on Zooms with them all day long? I'm on Zooms of them all day long. Yeah, effectively. So there is a CISO advisory board of CISOs who of current Zoom.

clients and they've been very helpful. I'm more, you know, I'm being, I'm a consultant, I'm working directly with the CEO, I'm part of their little war room of people who are working on these issues. So real quick, go over what your criticisms of the company were and what you're working and changing now. So there's really kind of two classes of issues that Zoom has to deal with, right? One is the traditional infosec issue, which is they just didn't invest properly in having a security team that was finding and fixing these bugs quickly enough. And so that work. is about building up the teams, building up the capabilities. There's three penetration tests going on right now.

And so we're both trying to find bugs quickly, but then also structure the teams in the way that that becomes an ongoing project that is appropriately staffed for a company that's this important. The second class of issues is related to their incredible growth because, as you said, they went from something like 10 million meeting participants per day to around 300 million in a very short period of time.

and that underestimates the difficulty because it's not just the growth of the people, but there's this huge explosion in use cases, right? And that's like really one of the big things that they're dealing with now is that they're an enterprise IT company that had this model in their heads of how are people going to use our products

for internal staff meetings, meetings with customers, meetings with supportive people, and now you've got things like journalists, you know, let's just say theoretically, opening up Zoom bridges for coffee chats, right? And then granting permissions to random strangers on Twitter that in the past would have only been granted to people within that circle of trust. And so that's really a safety issue. And that's something I talk a lot

about at Stanford and teach a whole class on. And that's about now considering themselves partially a consumer company and building their systems and their user experience in a way a consumer company would, not just an enterprise company. That question of being an enterprise or a consumer company has come up with Zoom in the past. I think one of the theories about why Zoom is so successful is that it has treated its product like a consumer product, even though it has a very traditionally enterprise business. And if you're going to treat your product like a consumer product, why do you think that those kind of consumer use cases weren't anticipated before? I mean, they got into a lot of trouble around installing web servers on people's Macs before, right,

just to make things simpler. And they seem, it seems like this company has sort of walked into that gray area several times in the past. Right. I mean, the web server issue is an interesting one because part of what they're trying to do there is having auto update set you know they basically do install a service that would allow them to update in the background which every major company does so google's got one called kS fetch that you see on little snitch all the time uh if you have a mac and you know they messed that up which a number of companies have you know the part of the problem here is a lot of these consuming use cases just didn't exist three months ago right people weren't doing yoga classes online people weren't holding high

school graduations online right and so like how do you invite several hundred teenagers to a video chat where any single one of them can put whatever they want on video and do so safely. Like that's just not, whether or not they're a consumer company, that's just not something people are considering in October of 2019.

So, you know, that doesn't mean Zoom doesn't have a responsibility. They absolutely do. But I see that as like a different kind of thing than the core infosac of finding and fixing bugs, right? It really is like a different skill set. And it's something that, you know, a lot of people are going to have to kind of focus on now

is if these products are going to be used to be, core parts of people's lives and not just something that facilitates enterprise meetings. You really have to design them differently. So Zoom has also been part of a conversation about the degree to which all companies now face maybe unknown risks, given that a lot of their workforces are at home and their security systems were not built for everyone to sort of be working remotely. I would love to hear you tell me what some of those risks are and maybe what we should be doing. in our lives to protect ourselves. Yeah, you're right.

Like, this whole thing is incredibly risky to enterprise IT. You know, lots of companies have to take into account that people will be remote, but they very rarely consider people being remote forever, right? There's lots of companies that support VPNs and such, but they assume, maybe not even explicitly, but they assume the way they set things up, that people are going to touch down every 30 days or 60 days. And we're now pushing, like, the 60-70-day mark.

And so this is 60 days without talking. to internal patching systems without talking to internal log aggregation systems, not talking to DLP, not talking to DLP, not talking to license management. People's active directory passwords are going to start timing out. If they haven't talked to a domain controller, they're not going to be able to rotate them. There's all kinds of crazy stuff. And a couple of companies have built zero trust networking in which your laptop, whether it's in a Starbucks or at home or in the corporate network, is treated the same way. That's a very small handful of companies. There's a lot of people who talk about zero trust, but they never go all the way like Google did. And those folks

are really in a difficult place because bootstrapping that kind of stuff once people are gone is extremely hard. And I actually have some other folks I've been helping out who just have kind of this normal enterprise IT problem. And it's significant. And it's really opening the door to attackers, which I think actually initially COVID was good for American companies. I was actually involved in helping a company with an incident involving some Chinese hackers, and they just kind of evaporated in mid-January. And everybody was like, whoa, this is interesting. Well, okay, let's fix things.

And looking back, it's because these guys were sent home, right? Like, the People's Liberation Army probably closed their office, right? But they're back. Like, they're back in the office. And now they know that everybody's at home, and they're outside of the reach of the IDSIDP. They're outside of the reach of the DLP system. It's quite possible their local logs are not being aggregated appropriately unless people have thought ahead. And so it's definitely hunting time for people who are working from home right now.

That kind of brings me to the bigger Zoom question, which is kind of related to how we talk about the scale of Facebooks and Googles. Is it better for Zoom to just emerge as sort of the big default video conferencing platform, have all the energy, have all the security consultants in the world talking to the CEO? is it better or is it better to have multiple competing firms with potentially varied attack surfaces? Because that is always the argument that Facebook makes about itself. Like, we will secure America, just give the whole thing to us, right? And it feels like Zoom's in that same kind of moment. Right, but Zoom is the upstart here, right? So who are their competitors? Microsoft, Google, and Cisco, right? All much larger companies with large, you know, who are trying to dominate their space. Cisco is as dominant network. You know, we don't, you guys talk, don't talk about

so much because very few consumers use Cisco products, but they are incredibly dominant networking. Google and Facebook are very dominant in their spaces. So Zoom is actually the little upstart. And this is one way you can look at their security problems is this is the cost of having a competitive space, right? There's a bunch of enterprise IT spaces that are highly competitive in which there are these smaller, you know, some of them public, but like mid-sized enterprises who are highly competitive against the big giants. And they don't have a thousand people on their security team like a Google or Microsoft does. And so the flip-sum is, is that unless something like this happens, they're not going to staff up to that size.

So I actually reject the entire premise, right? Like, the truth is, like, the Zoom being popular demonstrates that this is a space in which there actually is competition, that it hasn't been totally locked up by Google and Microsoft. I think it's fair to reject that premise. But at the same time, Zoom has been around for like 11 years. It's a public company. It has billions in revenue. theoretically when it does sell the enterprise security reviews are done. It just feels like

kind of the answer from Zoom has been, these are innocent mistakes. We're a young company growing fast. And I look at it, I'm like, you've been around for a long time and you have enterprise clients. How are we still working around the edges of what end-to-end encryption means? Right. How do you end up in that position? Well, I'm not saying that they, you know, the CEO has admitted that they didn't invest in security correctly and has apologized from the mistakes, right? Like, clearly all this investment is for that. That being said, like, if you look up the CVEE database on Cisco WebEx, it has a spectacularly long list of vulnerabilities. Just last week, Microsoft Teams had a bug that if you were sent a GIF that was

malformed, they could take over your entire enterprise account. That's an amazing bug. And you guys in the media didn't cover it at all, right? Because it didn't fit the kind of Zoom narrative. We covered it up. We covered it up to protect Microsoft. there I said it. I admit it, Alex. I heard Dr. Judy told me that. It's part of the same. Fauci works for Microsoft.

It's all part of the same conspiracy. No, but I'm serious. Like, enterprise software has bugs, including other products that are directly competitive. And it's right. Like, Zoom had these bugs, but you guys, like, the media collectively decides, like,

this is the narrative. And then all of the pieces of data are used to fit that. and then the parts that don't fit it just kind of disappear. And that just gets really frustrating from those of us in tech, right? Because it just feel like there's seven narratives that can exist on tech meme at any moment. And everything that's not within those seven narratives just doesn't exist. I will tell you that the Zoom and Slack PR teams are not shy about letting us know when there is a bug in Microsoft teams.

I mean, Tom is all of that stuff, Tom Warner, our team. But I think what you're seeing is it doesn't hit the escape velocity of, a narrative across the whole. Right. It's called Zoom bombing for a reason. Right, but Zoom bombing has nothing to do with that, right?

Like the people disrupting meetings is all about providing tools to people to be able to do meeting securely. In any situation, on any of these platforms, if you let strangers in, you run the risk of bad things happening, right?

And part of, like, one of the reasons Zoom created this problem for themselves is Eric announced very early that they're going to provide free accounts for education. And so you have, had in those first weeks when nobody knew what the hell was going on, before school districts

or schools or universities could figure this out from a central IT perspective, you had individual teachers going and just creating Zoom accounts and then getting them blessed as education accounts. And so Mrs. Smith would go and then post on the school website, here's the Zoom for our English class. And then there's guys out there who are scraping the web. And then they show up on Twitter and Instagram and Discord. There's these Discord channels where people are just doing Zoom raids. and they go share it with folks and then a bunch of people jump in and do nasty stuff.

That's going to be the problem for any platform that is successful. It's just like Zoom created the situation by, you know, especially in the education market of making it really easy

for people to sell. And that's the kind of the thing you see now is it's shaking out is now the school districts have followed behind. They've gone real enterprise accounts. They've put all their people on single sign on.

They've set all the settings correctly. And so it's become less of a pressing issue just because people have kind of figured it out. And then Zoom has changed things. of putting all these little switches that, you know, IT people will find or things that are actually callable by API of putting them in single places for people.

But that's the thing. That whole issue has nothing to do with whether or not the installer allows for a local privilege escalation on OS10. Right. Right. Or whether something's ended end-incryptor or not. And so, you know, it's that kind of conflation of these two totally different issues that I think makes things,

you know, a little more murky when we talk about it. Okay, so one last question about Zoom, and you actually wrote a long editorial about social networks in Beijing, and Zoom obviously routes a lot of, or did for a minute, rout some calls through China. They have a lot of Chinese engineers. How should people think about that relationship? Right. And so there's, you know, one of the things that I think they've really focused on is making it easier for people to choose whether or not you can use Chinese infrastructure, right? So one thing Zoom allows you to do is people in China can dial in two bridges. And so that means there are a scenario. in which you can have Chinese servers who are involved in that discussion. You know, if you're in North America, your client should not connect. And so there was a bug.

There's also, there's going to be more published about that specific study and how it wasn't really reflective of any scenario people actually normally run into. But one of the things that's happened is you can now, as an enterprise, go and you can unclick and say, like, these are regions where I never want people to be able to join. And if people there want to join, they have to basically come into the United States over the public internet and join. you know, lots of companies have Chinese engineers. I think it is absolutely reasonable to see a risk of people with those ties being leaned on by the Ministry of Security or the People's Security Bureau or the People's Liberation Army.

That is absolutely a risk. You know, from Zoom's perspective, they're putting those controls in place to reduce that risk. And I think one of the things that we're working on is, you know, end-end encryption, which would mean you wouldn't have to care about the servers at all, right? And if you had a true end-end encrypted meeting, that doesn't really matter if somebody's broken to the server, if they control it or whatever. As long as the client software was written correctly, you shouldn't have a confidentiality problem. Well, personally, I'm hoping that the Chinese engineers listen in on my Zoom calls because I want to talk to them about the virtues of free speech and democracy. And, you know, it's very hard for us to get into China in any other way. So I'll just put that out there as an invitation.

I like this sort of like Casey Zoom Air America idea where we're just going to like infiltrate. Trade the People's Liberation Army with hacks, zooms about how great free speech is. That's right. You know, we actually used to, you know, burner laptops from Facebooks that went to China. We would donate them. And I always used to imagine that there was this huge building, like this big War Games wall, full of like the cameras of all the machines that had been back to order coming back from China and that they were looking at all these like fourth graders picking their noses on these donated

Chromebooks and Macs that we had donated. That's amazing. So yes, it's one strategy. I said one more question on Zoom, but you just brought up end-end encryption. I've had a couple people tell me that it's almost too hard to build a Zoom that is fully end-to-end encrypted because oftentimes you want other people in it. You want to hit record. You want to do all that stuff.

Is that architecturally too hard or other ways to do it? So that sounds like a dare. You're right. I mean, so nobody else has done it, right? Like none of the enterprise video conferencing solutions are really end-to-end. I think there's two things here. So first, what was announced in the blog this week is that Zoom will be shamed. shipping an option for paid accounts where you have ended

encryption and a bunch of stuff's turned off, right? So you won't be able to dial in with a, you know, a normal 1974 AT&T Bakelite phone. You won't be able to dial in with like a polycom room system. You can use the Zoom room systems because Zoom controls those code, but not like the older SIP H3233 systems. And no cloud recording and stuff like that. So that will be the first version.

Where this really gets interesting is building architecture that then fails gracefully, right? So what we'd like to see is an architecture where then if you call in with a PSTN, you're getting most of the security benefits, but there is an exposure because now people are, but there are certain guarantees that can still be there. And this is about doing like practical security engineering and cryptographic engineering. This is one of the problems we have in the security industry is people are all about perfection, right, especially in cryptography, which really attracts these mathematically minded people

who are really good at criticizing and never building anything. and so, well, it's just true, right? This is like how cryptography works. Look, and there's a place for criticism, but there's also a place for building, like, practical systems that make compromises that actually keep people safer in the real world because those compromises allow the products to work.

And that's what's going to have to be kind of the challenge here is coming up with those cryptographic compromises and then defending them against public criticism of, well, if you want people to call in with a 1974 bake-light phone, then this is the best you can do, but it's still way better than the alternatives. Right. Okay.

So that's Zoom, which, you know, is sort of, I would say it's an expression of the pandemic, right? The thing happened, everyone went home, a lot of focus on Zoom because everyone's using it. There's a lot of those in the world. But let's talk about sort of the pandemic itself and now something I know that you're interested in cases very interested in, which is just the rise of conspiracy theories. I think most notably this pandemic video, which, has gone across every social network. One of the things I remember from our last conversation,

and the stakes of this were so much lower than the stakes of the pandemic misinformation, but we were talking about that Facebook copy and paste image of like, I don't give you a license for this, for my photos. Right. The Treaty of Rome. Yeah, right. Like every couple, like twice a year or something, like this thing goes around. And you said if Facebook wanted to do an image hash of that image and make sure,

it never appears they could, but they won't do it because they're sensitive about it. It seems like the pandemic has given the social networks a lot of clarity on what is right and what is wrong in a way that's easier than, I don't know, immigration or something. Why aren't they just stopping it at the beginning? Can they even do that? Yeah, it's a good question. So, you know, for video, a lot of this work happened after the Christchurch shooting, that the video that the Christchurch shooter first streamed and then was reposted over and over

again by his friends on 8chan. That triggered the companies to one strengthen the collaboration between them and this organization called GIFCT, the global internet forum to counterterrorism. So anyway, I'm not a big fan of the acronym, but

they did a lot of work to kind of be able to coordinate together, but it also, each of the company spent a bunch of time working on their ability to hash video. And it does not look like that kind of reaction is happening for the pandemic video. And I'm not totally sure why. I mean, there is a, I mean, this really is a difficult policy space.

And now, the pandemic video is so out there, like it says really, really crazy stuff. So I don't think you have to really argue about whether it's true or not. This is just like obviously. But there's a bunch of other kind of corner cases that are actually kind of hard of who do you trust, like the World Health Organization is one of the orgs that the companies are trusting and who's been wrong a bunch of times. Like just a month ago, they're telling us not to wear masks, right? Like, you know, let's not forget that like these are organizations run by humans but are also political organizations that are responsive to kind of their political masters. And for who that includes, you know, the PRC government, right? And so I do think there's some difficult policy decisions, but once they decided that pandemic is going to come down, their enforcement has not been great.

And I think they're not using the same level of aggressiveness that they did after Christchurch. And I'm wondering why. And one theory I have is that if you do that level of hashing, where, if a slight sliver of any video shows up in another video, that downstream video is blocked, then that's going to catch a bunch of legitimate stuff right now, because that pandemic video is actually playing. I've seen portions of it on MSNBC, right?

Like, you know, actual TV news is covering it. Unlike the shooting, the Christchurch shooting, people were exerting the video for like the first 24 hours, and then TV news stopped. It was mostly the Murdoch stations, right? It was like Sky and some people like that were showing like nonviolent components of it, but the people are like, wow, this is so disgusting. We're not going to show any of it.

That is not true with the pandemic where people are still exerting it. So I think one of the issues is that there's a lot of criticism of the video that also uses components of it. And their systems are set up to like completely nuke it or not, not to have that kind of, you know, call based upon, oh, is this criticism or not? So also, like, I'm actively reporting on this right now. I've, I'm talking to people from Facebook and YouTube right now to try to answer this exact question. I think that generally during the pandemic, they have displayed a much more interventionist stance when it comes to this kind of misinformation,

and they've been good at stopping these things from going viral, right? Like, I'm not somebody who cares if misinformation is posted on the social network. I just don't want it to get 8 million views if it's telling people that masks activate viruses, which pandemic says. At least in the case of Facebook,

it does seem like the initial move was instead of blind. locking it to downrank it in feeds, which is to me the least satisfying option, because I don't really know what downranking it means, and it still means some people see it, and who knows how that is being chosen. So anyway, I'm hoping that we'll kind of learn more, but it is curious to me that folks haven't nipped it in the butt. The one other thing that I'll say on the YouTube side is that after YouTube got a lot of criticism for letting various terrible things go viral. They actually built a team whose whole job is to see terrible stuff going viral and stop it

in its tracks. And generally speaking, they've been very successful, you know, at least compared to where they were a year ago, that team did not manage to nab this one. So that's kind of another curious piece of the pandemic story. So I actually, I think one of this demonstrates is that just taking the video down is perhaps not the right choice. Right. So the difference in Christchurch is the Christchurch is the Christchurch video was one, it was a snuff film of people being killed. It was also the existence of the video was was and then this manifesto that went along with it was meant to cause race war right and so I think you can argue strongly that that thing should just never be

multiplied or exist just like child exploitation doesn't exist the whole point of the pandemic video is to make people believe there's a vast conspiracy of people trying to suppress this information I don't think the best way to beat that is then to suppress the information right like I think this is actually a situation we're downranking and then especially labeling everything with, hey, this is not true. Everything you're about to see is not true would be a better response because they knew it was going to get banned, right? They absolutely, one, so my colleague Renee DeResta has done a bunch of work on this. The whole, this whole process is totally commercial,

right? Everything behind Judy McEvitz, these are people who are doing this as like grifting, right? They're doing this to make money. And so there's a whole set of professional grifters who are helping supporter and they knew this was going to happen. One of those grifters, is ex-Google. And so they have set everything up. And then the companies were effectively forced by the media coverage into doing what they wanted, which was suppress it, which then gives them incredible power on the video. I think this is a great example of like if you leave the official one up, but it almost never shows up in people's searches and then you make people click through a warning, that that would have been a smarter response, even if it's less

satisfying. I'm sympathetic to that argument if, you know, if what you're trying to optimize for is the fewest people believing that putting on a mask is going to give them the virus, right, which is like very counterproductive to public health. I think where it gets tricky is we still don't fully understand how these sorts of videos benefit from algorithmic promotion, right? And often these platforms are used to essentially recruit customers for these terrible ideologies and are often very good at it, even if unwittingly. And so I think the question is, if you do leave a video up on a platform, whether it's Facebook, YouTube, Twitter, what are the algorithms, the recommendation algorithms that run on those networks do to find other

people who then become true believers and super spreaders? But that's the whole point of downranking is that it should never show up in any recommendations. And then I'm going to be honest, like the algorithmic thing is the nice little out the media takes for themselves to not address like the core free expression issue here, which is the core issue here is that if people are deciding they want to be part of groups that trading these conspiracy theories and they decide that they want to see this video. An adult has said, I want to see this. Then I think we have to have a very high bar in a democratic society to use centralized control by trillion-dollar companies to prevent that. Right. And in doing the prevention, it creates a

legitimate argument from their side that everybody's aligned against them. And then that feeds back into these grifters making a ton of money. So I just, I just feel like the algorithmic thing, It's the thing people throw out instead of like talking about the hard tradeoff. It's like, oh, well, it's the algorithms. Actually, no. What's happening here is that there's people who want to see it. And then there's a large number of people who want to provide them with the video. And when you have a huge amount of demand, a huge amount of supply, what you've got in the middle is the companies putting their fingers in the dike.

And then putting their fingers in the dike is what is powering the conspiracy. Yeah. I feel like maybe the best way to fight the conspiracy is to reveal to people that the architects of the conspiracy are not like in a beautiful conference room and a, like, all black building, but it's an ex-Google engineer with a MacBook and a Shopify account waiting to sell you supplements. Yeah. Like, as the conspiracy stuff continues, that I'll be honest, there's a part of it where, like, I'm comforted by the idea of a bunch of intellectual elites in control of corporations with

a plan because it seems preferable to our current reality of no plan whatsoever. And I think that's like, when you say the free expression issue, people are choosing a fantasy over reality, right? And the platforms are enabling the choice to just buy the fantasy whenever you want, as opposed to contending with the reality. That is like a deep philosophical issue for the platforms. But just putting up warning screen saying, this is not true. They are private companies. They could just say this doesn't exist in our space.

Right. And like where in sort of in your past, like where did you find those lines being drawn as the platforms would say, I mean, Facebook's oversight board is like the purest example of this. They so want to be the place where free expression happens. They've set up a court, like mimicking what most countries would do, instead of just saying we're a private company, we can make decisions. I mean, I think the pandemic and then a lot of things post-Trump's election have made me rethink what I always thought would be true, which is if you help people understand where disinformation comes from,

they won't believe it. But it turns out that the kind of attraction to the tribal aspect is, even if you know you're being played, being played makes you feel good, and so you'll do it anyway, right? So that is something that I was probably wrong about, and that underlied a lot of our work at Facebook, which is like, okay, well, you know, maybe we're not going to suppress the speech. We're going to make it super clear that this is paid by these people and this is this and this is that. And, you know, unveil, pull back the curtain so you can see the man behind it. And it turns out people are like, oh, that's fine, right?

It's like, because there's this vast conspiracy on the other side. This specific one, I feel like this might be like a self-correcting one though, because the truth is, is like the people who are going and going to Walmart without a mask and are going to protest and stuff, like they're going to get sick and their friends are going to get sick. Like if the three of us like coastal globalists on this call, if we believe the science is true, which I do, right? I've been locked away for a couple months. I have like a history of one issue, so I'm actually really afraid of this virus. I'm, you know, I'm 41, so normally I'd be okay, but I'm not like in the great category of people of 41-year-olds, it turned out well.

So, you know, we're locked away and we've been doing so before it was legally required. If we believe the science, then this will be something that people are going to learn the hard, tragic way that they've been fooled. And so I, I, there is a part of me that wonders whether or not this becomes a moment where these people who are willing, because like the impact of electing Trump or believing this stuff or believing that, has been so disconnected from them. You know, the pain has been felt by other people that the fact that their grandparents and their parents are the ones who are actually going to suffer for them being willing to be fooled

maybe is a turning point. Maybe not. Who knows? When you look at the sort of other related conspiracy theories, there's a video of the people in Australia chanting arrest Bill Gates over the weekend. Every day we get a note about another 5G cell tower in Canada being lit on fire or something. Yeah.

Does that stuff seem coordinated? Does that seem like it's being egged on by foreign governments? Does it seem like it's the same sort of grifters at the heart of it? Or is that more organic, do you think? So, I mean, the propaganda space that we've seen around COVID is actually super complicated. You know, for the first time, really, in history, every government in the world has the same number one concern, right? Maybe since a world war. But, you know, in our lifetimes, this is the only time that every government in the world has the same number one concern, which is defending their own response to coronavirus.

and then in some cases trying to hurt others through their response. And so what we've seen is a shift of effectively all of these pre-existing disinformation actors have all shifted to COVID as their number one topic. And in some cases, it's because they're trying to deliver a message around COVID. So like the Chinese, right? Like they want to basically deliver the message that they didn't cover anything up, that China was responsible. China warned the rest of the world that everybody else ignored them.

Whereas then other actors, like Russian actors in Eastern Europe, it's just reactive of like, oh, well, here's the number one topic. And so their overall goal to talk down the EU in the Polish elections, COVID is now the topic they're going to use, but they don't really care about it. And so it's hard to disambiguate because it's effectively all of the disinformation activity has shifted this way because it's where everybody's attention is shifted. And in the U.S., a lot of it's unaligned, right? So you have what you have to call official propaganda coming from the White House, right?

But then it is all these private organizations that are either, you know, you know, like the people behind Plannedemic are either kind of hiding or people who are very obvious, like the O-A-N-N, right, like, which is a news network that is carried on DirecTV and AT&T that has a seat in the Washington, you know, in the White House press room, is pushing the same kind of things. And so I think it's like kind of an open conspiracy is what you would call it, of these people seeing it within their economic best interest, but also their political best interest to push these ideas. Support for this show comes from Shopify. Every thriving successful business has to start somewhere.

A good place to start is a relatively simple question. What if, given the right tools, I really put my all into this. One tool that can help grow your sprouting business to new heights is Shopify. Millions of businesses around the world rely on Shopify for e-commerce. They offer a host of helpful tools you can take advantage of, from payment processing to analytics to website. design. Their design studio includes hundreds of templates to help you create the exact website you've been envisioning for your business. If you're wondering, what if I need help, then no worries

because you're never left to fend for yourself. Shopify's award-winning customer support is available 24-7. It's time to turn those what-ifs into a thriving business with Shopify today. Sign up for your $1 per month trial today at Shopify.com slash vergecast. Go to shopper.com. Go to Shopify.com slash Vergecast. That's Shopify.com slash vergecast. Support for the show comes from Upwork. The days of doing it all,

all by yourself, are over. There's no romance and burning out while you're trying to scale. Instead, you can check out Upwork. Upwork helps grow your business by giving you fast access to specialize talent across more than 125 categories so you can fill skill gaps,

launch projects faster and scale without committing to full-time headcount. And finding the right talent is easy. You can browse profiles, review past work, and get help scoping the role so you can get started quickly. Seriously, you could connect with the right freelancer in just a few hours, especially when you sign up with Business Plus. Their AI-powered shortlisting pairs you with the top 1% of talent in under six hours. No endless searcher required. You can visit upwork.com right now to post your job for free.

That's upwork.com to connect with top talent ready to help your business grow. That's upw-w-rk.com. Upwork.com. One of the fears last year when Facebook announced that it was going to move to private messaging is that it was going to take a bunch of bad behavior that was previously viewable to all. and take it underground where it would be harder to see and shed light on and provide correctives to. Is Plandemic the first real case in which we have seen that borne out? And how much do you worry about that as just kind of a force in our discourse?

You know, it's a good question of how much the private conversation is driving Pandemic. I don't think we have any good empirical evidence of that. Certainly a lot of the places it's being posted are in private groups, right? So you have the really private messaging of like small, small groups. But then you have this big mushy middle space of like Facebook groups that maybe have a thousand people in it. And so they're not close friends, but they are people, they are screening for people that they think are like-minded. And that's like an interesting space of what is the responsibility of the platforms from that. This is one of the fundamental hard tradeoffs.

If you believe in people's privacy, if you believe that people have the ability at private conversations, then you also have to believe that they have the ability to privately talk about that. things that are not true, right? And in some cases, harmful. One of the nice things is that this is, again, a self-correcting problem. As people move to smaller spaces, they get less amplification, right? So if you're speaking in a Facebook group of 1,000 people, you're limited to 1,000 people seeing it versus a public post that a million people can see. So, you know, I think we have to carefully calibrate what level of work do we want the platforms to do in different spaces. And I would say when it starts to get into private groups that people have curated themselves, and then again,

adults are choosing they want to be in that space. We'd be real careful. There's a series of articles by Brandy and NBC about a woman who made a tragic decision around her pregnancy that ended up with her baby dying. And I was very conflicted about that because the truth is, if you believe in reproductive freedom, if you believe that women have a choice to not go to the doctor to have birth on their own, then they also have the right to talk about it. And so if somebody has chosen to join a group where people are supporting each other in these life choices, then I think we've got to be real. careful about saying that, again, like a half trillion to trillion dollar company should go and squash that, right? And so I do worry about things going there. And I also think it's a flip

side of some of my beliefs on privacy that we have to allow people to be wrong in private. I just wonder about like what kind of pipelines are either exist now or will be built between the private messaging, you know, the WhatsApps and the signals and the private Facebook groups to those public groups, right? Like do they become sort of these new breeding rads? Like all of this is still kind of circling around the question we kicked off the podcast with, which is why does pandemic have 8 million views? You know, it may be that it was all just folks chatting in these public Facebook groups and it kind of took off from there. But, you know, I wouldn't be surprised if it actually got its start in private messaging. Yeah, that's true. And the model for that

would be like Christchurch, right, where 8chan was effectively the small board where we have a couple thousand either trolls or white supremacists or both who were coordinating then to get the stuff posted in all kinds of other places. And so, yeah, right, like that there is a model there of, you know, you can have 100 people in a private WhatsApp message who are then reaching millions of people because in those hundred people are able to coordinate how they're going to get the video out, get the whatever manifesto, get their information out. Can a company like Facebook see that activity, not inside the group, but the sort of link

to the activity within the group to what happens next? Possibly. I mean, certainly if they can see inside the group, they can see those people linking out. you know, we always had a higher predicate of suspicion to go and look at people's private messages or locked groups, right? Like public Facebook groups are available on CrowdTangle. You can just get all this data, you know, from people's public postings. But all the private stuff is not. And there's a pretty strong privacy argument for that.

And even inside of Facebook, we had a higher level, you had to basically have a higher level suspicion if you're going to break that open. And so in those cases, yes, I've seen like, is there a conspiracy among people like on a totally different site or, on telegram or even on WhatsApp, that's a lot tougher. In some cases, you can see it because you can see inbound links. And that's one of the things that I did congressional testimony last year. And those are the kinds of things I actually would like to see the companies be more aggressive of is trying to understand where refers are coming from and whether or not, you know, and coordinating with each other, but then also infiltrating. And, you know, the example of that working out was ISIS, right? So by, you know, the end of kind of ISIS being effective force online,

we had full-time people on Facebook who were sitting on telegram channels who were watching ISIS discuss what their plans were and then seeing like, okay, great, here's our new video. And so they'd grab the video and hash it before these guys had even gone it to the distributors.

And so you can get that aggressive. It's just like, you know, for a violent international terrorist group, people feel pretty comfortable about that. For white supremacists, people feel very comfortable. Having private companies like infiltrate anti-vax groups that gets a little trickier, especially if the underlying behavior

is not illegal, right? Which is the other issue here is that when you talk about terrorism, it's illegal. When you talk about child exploitation, it's illegal, being a jerk and like saying arrest Bill Gates is not illegal, at least not in the United States. I feel like Bill Gates, he's going to invest some money

towards making saying arrest Bill Gates illegal. He is, by the way, the funniest part of this. I know Casey's talked to him, I've talked to him. I interviewed the guy who made the inside Bill Gates. he is so exactly who he says on the tin that the fact that he's the focus of conspiracy theories is at once like tragic and sad but very deeply

funny. Well like what's the upside for him other than like what does he get? He's already like the world's riches fan depending on what like the stock market's doing right? So like what what's like the secret thing he wants? Like the guy to do whatever he wants and he's decided to like wipe out disease with

the rest of his time on this planet and like where's the side gig on this? Compared to the grifters who have an obvious economic motive of pushing something like pandemic. It's just like it's kind of hilarious. But also, you know, it shows you again how the tribalism fits in, right? And I think this is actually what I'm really fair for the election is, you know, I don't subscribe to Donald Trump's a secret genius theory,

but he does have an incredible capability. You put his finger on the pulse of things that people like us, again, us coastal globalist elites of people who have made United 1K at some point, right? That we do not see and that he does. And I think he's done that with this of like putting his finger on the pulse of people being angry about the shutdowns and looking for somebody to blame. And, you know, this might be a really effective force going into the election for him to dodge the responsibility that he clearly, clearly had. And to push this on to other people. And I really am afraid that this, that we're underestimating how much pull there is.

Like, again, we can talk about algorithms and stuff. But there's clearly like an underlying set of beliefs here that people have that are making them. really want to believe these kinds of ideas. This actually brings me to the thing that theoretically you're here to talk about, which is your Washington Post op-ed. Obviously in 2016, you know, the Chinese and Russian governments were very good at playing into Trump, whether or not you believe there was collusion.

But it was like very clear that there were ops and social media platforms designed to sow division. Trump is very good at division. You just wrote that social media platforms in this country are at risk of being taken advantage, in particular, walk me through that argument and walk me through the issues you're saying. Yeah. So, I mean, we've made a lot of arguments in lots of places. So it'd be clear, you know, I also think there's things they need to do differently around the domestic disinformation and around transparency and political ads. This specific op-ed, though, was about state media.

And we use the example of Chinese state media in that, you know, the Chinese state is trying to push these ideas around COVID that are not true and are doing so effectively through advertising specifically on YouTube and Facebook. They are being countered by some non-state actors like the Epoch Times. And so like there's this fascinating war going on on YouTube right now where the Epoch Times which is run by Falun Gong,

the religious movement slash cult, which apparently has a lot of money because there's no way that these YouTube ads are profitable are running all of this anti-China anti-Trump stuff at the same time that then Chinese state media is advertising there. So on the state media side,

something that we proposed in our op-ed was that we think that there needs to be a rule of the platforms to treat state media differently based upon whether they're coming from autocracies or not, right? So, you know, there's a difference, clearly, between CCTV and Shinhua and PBS and BBC, right? And that the standard that we proposed is if you're related to a state that blocks that social network

in your own country, then you should be treated differently. At a minimum, you should not be allowed to run ads. Perhaps you're not allowed access to a bunch things like verified accounts and the like. Because our theory being, if the BBC wants to put out a line, you know, I think they're editorially independent, but let's say they weren't even that independent. At least in the UK, people can use Facebook to push back against that, right? You can comment safely on a BBC story. You can say things about Boris Johnson. You can't do that in China, right? The Chinese people do not have access to open social media to criticize their own government.

And so in a situation like that, giving them the ability to pay to amplify their message in the West while suppressing speech domestically is just an asymmetry we should grant. That being said, that doesn't deal with the Epoch Times issue, which on that side, I think the big problem is due to some criticism from the media, YouTube created this category

of political ads that come from media companies don't have to be included in our transparency, and they need to give that up because there's effectively no way that you can say that Epoch Times is not media, but Voxes, right? Like, you know, they check all of the boxes there's no box really you're run by a cult,

which I think the verge probably is, right? So even in that case, we don't want to make that a standard. It's a net neutrality cult. But I think, like, COVID is burning up a bunch of issues that, like, gaps that need to be closed and hopefully closed before the election. Do you see the platform's making moves on that stuff? The China stuff's hard because they make a lot of money in China,

even though they're blocked, right? Like, there's a lot of Chinese companies that advertise on Facebook to the diaspora and to Western consumers. And so there is some money in play here of possibly having that. I think also they never really want to confront the big states directly. Right. And China is, I've said this before, China is the ethical black hole for Silicon Valley,

which is all of this stuff the companies are willing to do for Russia. They will not do for China. And it's because Russia is economically irrelevant. Nobody cares about the Russian market. There's no Russian supply chains don't work, but they'll do it for China. Right. So Apple will do things for China. won't do it for anybody else.

Facebook will do things for China. They won't do it for anybody else. And I think like the PRC is the place where all of these kind of ethical frameworks fall apart because of their economic importance. So just while we're talking today that, you know, Trump is tweeting that the FCC should fire Chuck Todd. Yeah, is that how that works?

It's crazy. It's not how that works at all. But it's like obviously a pressure campaign about speech. We were just talking about FCC commissioners over the weekend. Brendan Carr, who's an FCC commissioner, did a lengthy tweet story. of how the Facebook Oversight Board is composed of these leftists who hate Trump. Does that pressure work on the platforms? Does it affect those kinds of decisions about what gets oppressed,

what gets taken down? How do we treat China? Oh, absolutely. I think that's actually one of the lesser understood stories of the last couple of years is how the Trump administration has effectively used the levers of power to pressure the platforms. And in fact, a lot of the kind the anti-tech media has played into that by amplifying their thing. So there's a great example of this that was obvious to those those who weren't in tech, and I think went by people's radars, which is Department of Health and Human Services filed a lawsuit under the Fair Housing Act. And in my understanding, it was like the first time HHS had done this under Donald Trump.

Who was the target for the Fair Housing Act? Was it some slum lord? Was it some, you know, county that wasn't appropriately setting up the county housing? No, it was Twitter and Facebook, right? This was obviously because they were trying to do a brushback pitch of we control the regulatory state. Our ability to make your lives hell is effectively absolute. And actually when it got very little coverage, but those who covered it actually got covered as kind of like, yeah, the tech guys were really bad about allowing these bad housing ads, right? And not seeing like there's a reason why these are the only people.

And so the ability of the administration to say you're monopolist, you're, you know, you have antitrust issues, you have privacy issues, and then to turn all of that into, you know, public arguments that then in the back end, what they're basically just asking for is for their content not to be censored, I think is incredibly strong. And Democrats just walked into this, right? Like Elizabeth Warren was saying things that Ted Cruz was agreeing with. And I'm like, that should be like a life warning. If Ted Cruz retweets you in a positive manner, you really should like take a moment and, think about how you got to that place. And it's not because Ted Cruz is actually, you know, it's not Josh Hawley doesn't really care about monopolies. Like, you know, Tom Cotton doesn't really care about these things. It's like they are throwing a brushback pitch of we control regulation. We control prosecutors. We can make your lives hell.

And I think that's absolutely had an impact on content moderation decisions. I think Josh Hawley does truly, deeply, sincerely hate Google. I mean, like, there's one thing, you know, we've encountered him several times. Um, the hatred for Google in particular seems like a consistent and sincere position. And where does that come from, though? I don't know. He just, maybe he just like got a, maybe he like bought something bad on Google shopping one day. But like, it's just like very clear that he, his antipathy, him in particular, his

antipathy for Google, not, maybe not the other ones, but Google in particular seems sincere in a way that I think is super interesting. Like, I think it's a mistake to cast this all on partisan lines. Like, I think Elizabeth Warren's critique of tech dominance. is also sincere. They come at it from ideologically different places. But I don't think that either one of their arguments has anything to do with Trump tweeting at Al Jit Pai

to fire Chuck Todd. Like, that's crazy. I think those effects are very different. But I do think, I mean, my point here being Elizabeth Warren and Ted Cruz Campo be right, right? Like, you can't say the best way to get the content moderation I want is to break these companies up and to say so from both sides. One of them has to be wrong.

And then also, like, Democrats just, haven't been smart, I think, about making sure that they frame their complaints in a way that is actually productive. And so if all you're doing is saying that tech companies are out of control and Section 230 gives them too much power, then Bill Barr can pick that up and turn it into great. I'm going to use it for exactly what I want, which is I want to outlaw encryption, right? And that's effectively what has happened is that without having like really, really crisp specific recommendations of this is what we want to happen. If you just have these kind of general feelings of I feel that this is wrong, then it's really easy for people who actually have

their hands on the power to then utilize that in a way that they've always planned on. Sorry, I think Casey wanted to jump in. I'm sorry. Well, I mean, I was going to say that the reason that both Ted Cruz and Elizabeth Warren were, like, complaining about the tech companies is because, like, finally something had grown to such enormous power that even the most, you know, diametrically opposed viewpoints could find some common ground, which is this thing, you know, needs regulation. And of course, they would go about it very differently. So, you know, I think the fact that there was actually some bipartisan agreement speaks to the magnitude of the issue. And of course, we've gotten very little

movement on any of that. But I don't know. I, you know, I basically agree with you that a lot of critiques aren't very nuanced. I still think Democratic critiques are generally sharper than Republican critiques, which to me mostly seem like they're in bad faith. Like if Josh Holly had to live in the world that Josh Holly has proposed where content moderation is illegal, I suspect he would find a whole bunch of new things to complain about that Google and Facebook and other folks were doing, right? You don't think Diamond and Silk have like a really good argument that needs to be backed by the U.S. Senate? Yeah, exactly. So, yeah, as always, we should be moving toward nuance in these discussions. Nuisance and then also of like productive suggestions, right? Like,

I do think there needs to be regulation too. What I don't like is when people just throw out these kind of broad critiques that then there are other people who have a plan and they're going to execute that plan. Right. And we see that. Like, one of the untold stories is that the other untold story is that lobbyists for the ISPs have been incredibly effective, right? Like the people, AT&T, Verizon, Comcast, these people are so plugged into DC. And so whenever somebody criticizes the people that they, they hate Google and Facebook because they don't like the idea of the money accruing on the edges of the internet, right? They want to be able to tax in the middle and they're super pissed about net neutrality. And so that's the other untold story here is that you've got those ISPs working with

the news media, you know, the news media alliance and the other, you know, content next, digital content next and the other people who represent kind of the old media. And they are collectively very effective at pushing these ideas inside of DC. It's funny. If you're listening to this and you're hearing these trade groups, Alex is mentioning, you might think that they're like secretive or on the edges. They are very loud. Like, none of this is actually that hidden.

Like the ISPs hating Google, DCN, which is run by Jason Kent, who we should actually just have on the show someday. If you want to have somebody who's paid like $900,000 by Rupert Murdoch to talk crap about tech, if you think that that's appropriate, sure, you should have them on. I mean, it's a spicy Twitter account. That's what I'll say. It's always interesting to talk to. But like, it's happening really, really loudly, right? And to Casey's point about nuance, there's not a lot of nuance in those tribal lines, which kind of brings me to the next big thing that's happening. There's about to be an election in this country. I doubt there's going to be a lot of nuance in either one of these campaigns. It doesn't seem like it already.

I would point out that Dan Suvino, Donald Trump's campaign manager, likened this campaign to the death star. So it just doesn't seem like there's going to be nuance at all. And then that does seem like a very ripe vector for misinformation for foreign interference. How are you seeing that begin to play out? I mean, my biggest fear around the election has always been attacks, from the foreign perspective, have been attacks against the legitimacy of the outcome, right? And we saw, you know, there's actually a great piece

that I think people should read by Franklin IV in The Atlantic today. It's a very long piece. I know he's been working on it for months. But it's a great summary of a lot of the issues that happened in 2016 in which are relevant to 2020. And, you know, I think he does a good job

of, you know, once again, highlighting, and again, this isn't a secret, but it's crazy when you see it on paper of how every part of the election protection infrastructure has been taken apart by the White House, right, and that we now have a handful of people, a couple of which are quoted, and I'm really afraid of their names being in there

because now the White House knows they exist. But there's a handful of people in the U.S. government who have been trying very hard to protect our elections against external interference and have gone no support from the administration, and I'm really afraid of them being sidetracked now. Anyway, it's a good piece. And my biggest fear has always been what I talked to him about,

which is, you know, the Russians dip their toes in attacking election infrastructure and then didn't end up following through in 2016. And in 2020, you know, my big fear was they would attack the election infrastructure to mess with the voting counts, to mess with the voting rolls, with the e-pull books, to cause chaos, and then to have a disinformation attack that would say, oh, well, this election's been stolen by whoever won, especially if Trump lost. And I don't think they need to do the hacking anymore.

Like, that's the thing that COVID has done is that the election is going to be chaos no matter what. You don't have to break into anything to make it totally. crazy that every state's going to have different standards here. And some are going to do vote by mail and some are not. And some might do internet vote by mail, which is nuts. And in some are, they're going to close a ton of the toll places. And so you'll have people six feet distanced wrapping around the block five times. That kind of stuff is going to create the underlying chaos that you need to question the election. And there is absolutely no way Donald Trump ever says that he legitimately lost. And it is in the best interest of a number

of foreign adversaries that if it goes that direction for them to support that idea. And maybe even on the other side, if Trump wins, then they support, you know, Biden saying that it was stolen. And I think, you know, the example here is the Iowa Democratic Caucus where there's no hacking. It was just us screwing things up. And domestic disinformation actors stepped in and started casting aspersions immediately. And they got some amplification overseas, but they didn't really need the help that much.

And so I am really afraid. And so my colleague Nate personally at Stanford does a lot of work around this and is a big proponent of vote by mail being the solution here. But now we've got Donald Trump saying that vote by mail is a scam, which is really even though he votes by mail, which I think is great. So what do you think the next steps are here? Are there any? Is there a way to to?

Unfortunately, it's too late, right? Like the truth is, is like, on, you know, when the new Congress was sworn in in 2017, that's when they should have started on election security. we need probably more centralization of our elections. There's around 10,000 election authorities in the United States. Just getting the list of all those people was a huge project by DHS. And we need to have at least state-level security teams that really know what they're doing. Right?

We can't have 10,000 teams, but we could have 50. And some of that has happened. And there's some states that have been really thoughtful. Colorado is a good example. California, Alex Bedia, has done some good work in this area. But for a bunch of the swing states, they haven't. Ohio being the exception. Ohio's done a bunch of work.

and it will be interesting to see what they do around COVID. But I think it's just too late. Like, you know, ballots have to be getting printed right now. Machines have to get being bought right now. And there's no money for it. There was no money in the last stimulus. I mean, I guess, you know, the last thing could be huge grants for the states to be able to support this kind of stuff in the next stimulus bill. And I know a number of Democrats are pushing for that.

The possibility that making it by Mitch McConnell, when the president has said that vote by mail is about election rigging, I think is very unlikely. Do you see a bigger path for the platforms to play a role this time than last time? Yeah, and certainly. I mean, I think as you read in the Atlantic article, it talks about the stuff we did before 2018. And 2018, you know, in some ways, was just kind of a dress rehearsal. So I know the companies are much more engaged. They're working with each other.

They're tied into FBI and DHS. You know, I think those are all good. I think the real question is if we have a Iowa caucus-like situation, what is the role for the platforms to play? You know, because one of the things that my colleague, Nate, points out, is that in a vote-by-mail situation, election night lasts for four days, right? Like, you know, it will take a week. And so you've got all these people glued to their TVs, and you've got the anchors up there vamping hour after hour, and you create this huge vacuum of answers. And you'll end up with the legitimate media organizations are not going to call it, right?

Even Fox is not going to call it that night. I don't think we're going to see a Carl Rove going down the hallway situation. You know, they're not going to call it until there's enough data in, and it will take multiple days to get that data in in a vote-by-mail situation. And so what happens during those three or four days, I think, is going to have huge impact on the future of our democracy, because it's in those four days that we decide of when they announce who the winner is, whether people accept it or not.

And so what should the companies do in those four days is a big question. And I'm not totally sure what they can do other than to try to, to drive down virality, drive down amplification. Going into that, they need to have done all the work possible to find networks of fake accounts and have them flushed out, right? So you don't want all of a sudden the day after election for 100,000 Twitter accounts or a thousand Facebook accounts to wake up and start pushing stuff. So they need to be doing that work. And then they're just going to have to have to have human beings that can react super quickly, right? They're going to have to be watching what's going viral, look for a pandemic like video. You could totally see

somebody cutting together a video with stock footage of like, oh, well, these are ballot. it's being stuffed, and I record this in the back room in California, and this is how it's being stolen. Like, people will prepare those things and then dump them the day after. And, you know, that's the kind of thing they're going to have to react very quickly to. But as we talked about with the pandemic video, the fact that these, you know, West Coast globalists are the ones who are suppressing it is going to be, you know, is going to feed into the conspiracy theories. And you're probably going to have, you could very possibly have the president amplifying that, you know, being covered by all the networks in real time, because he's doing it from the briefing podium of

the White House. So I really am afraid of this election, unless it's like super obvious, one-sided. If it's within a couple percentage points, it's going to be a really rough December in January, you know, until the Senate meets in January and verifies the electoral college, which fortunately the person in charge of that's Mike Pence, so I'm sure that will work out just fine. This brings me, it's just, I keep flashing back to the election of 2000 when the whole country is like, people in Florida can't use this complicated ballot.

it and that held the country on pause forever. And it just seems like this is going to be a thousand times more complex and it's in the shape of it. Right. And Al Gore saved the Republic. Right. I'm just going to say it. Right.

Like the fact that Al Gore was willing to jump on the grenade when he had all kinds of legal arguments he could have made and political arguments he could have made. And he was the president of the Senate. He could have gone in and made that hell. And that guy jumped on the grenade and took it for democracy. And so whatever you think about him and the other work he's done, you know, that is like the last kind of statesman-like behavior like that that we've seen from an American politician. Yeah. All right. I'm going to end with basically a game show segment in honor of Casey's late

great podcast. It's not as much of a game show. There are three major social networks in this country, Facebook, Twitter, YouTube. That's how I think of them. Instagram is part of Facebook, obviously, but Facebook. And Facebook is very clear that it's all one structure underneath. How are each of them doing right now in dealing with the various threats that are out there? You run the internet observatory, so I'm asking you to observe. Yeah, for those three. So Twitter has been the most aggressive in their policies. They always have implementation issues because they're the smallest of those three, right?

So they just have the least money they can throw at things. Their model of authenticity also makes things difficult that people are allowed to have. But I think they've been the most clear about what their policies are. And so I think you have to give them a lot of credit for basically, like, they have an opinion. I don't agree with all the decisions, but shit, they go out and they say it, right? They're like, this is what we're going to do. This is our opinion. We're going to enforce it.

And you know where they're stand. And I think they're pretty consistent. Facebook has by far the biggest team working on this. So they've invested the most money. I think the enforcement between Facebook and Instagram has been extremely uneven. Right. And so there continues to be kind of the Instagram people acting on their own. And I, you know, I know what it was like when I was there.

I would think that would have been fixed in the last couple of years. that's still like a significant issue is that there seems to be the significant gap in policy enforcement on Facebook versus Instagram. And when you talk about things that deal with young people, Facebook's not important. It's Instagram. Now, when it's the boomers, it's Facebook, right? And so I think they've invested the most and they've put themselves as the center of the coordination strategy. So all of the roads to do this coordination runs through Facebook, which is great. I think they've also been the ones who have been worked by the refs the best, right? So they're the ones. One of the things I don't like about Facebook, it's actually, Facebook in some ways is too receptive to media criticism. and that comes from both the left and right. And so in situations where people complain about something,

they will change their mind. And so that's something I hope the oversight board actually helps with is that when something's very difficult, they kick it to that process instead of, oh my God, people have written too many editorials. We're going to have to change our mind. And so I am worried about them kind of flip-flopping very quickly through it. I also don't like the way they've allowed the Chinese state media

to use the platform. And that's something I wrote a whole washing post. Google, Google. This is how it's almost a whole answer. Yeah. Google has some of the best people working on this. Their strategy since 2016 has been to quietly deal with issues

and to like Facebook take all the crap. And so they basically hide on everything. They are the hardest to work with from an outside perspective, right? So from a researcher's perspective, Twitter's the best because Twitter just dumps out all of this data publicly. Now Twitter effectively has no private network.

They have DMs, but they don't have private groups. So the privacy issues for them are not as sufficient as a big deal. Facebook has built all these APIs, but they've been very burned on privacy. So getting access to stuff requires special NDAs. It's always like a fight to get access to data. And they're very worried about European regulators in this. Google shares nothing.

So Google will go find something secretly. They will wipe it out. They will never tell anybody. And this is working incredibly effectively for them. Because everything you have read about the Internet Research Agency on Facebook, in 2016, if almost everything you've read, came from our team, from the team I supervised at Facebook, right? And then we gave it to the special counsel, and they ended up at the mole report, and we gave it to Congress, and they released a bunch of stuff publicly. Google had a ton

of stuff. They just secretly took care of it, made it disappear, nobody ever wrote anything about it, and then they just kind of quietly slunk by, and nobody talked about it, and spent two years beating up on the company that was public. The Google side makes me very upset because they have some incredibly good people in this area, but their overall corporate strategy of silence has been really effective. And if you're an outside company and you're looking at those three companies,

you're like, I want to be Google. I don't want to be Facebook or Twitter. And then the other companies that are kind of up and coming here are the big ones TikTok. And TikTok's starting to deal with this stuff. And they've got like, you know, some really legit trust and safety people who have joined who are thinking a lot about it.

But they're also having to work within the strictures of bite dance, a Chinese company. And they're having to bootstrap very quickly. they also have a difficult disinformation problem in that their content's mostly visual. And so that just makes technically this a lot harder than stuff that's tech space. Do you think TikTok will be potentially, I don't know about a decisive player in the election, but is it possible that that's like sort of the disinformation vector that surprises everyone this year? I think it could be, yeah.

When I think about this, I just model it out. Like let's say I work for GRU or I work for Progosion. And I've been asked, okay, I want to, one, drive divisions in American society, and two, I want to support Donald Trump again. How do I do that? And the best way to do it is actually kind of a replay of 2016, which the best way to actually support Trump's election is to peel off Bernie voters away from Biden, right? Like that's, that was the GRU playbook in 2016. It actually worked pretty well. And where are Bernie voters are on TikTok and Instagram? And so I would, if I was the Russians right now,

I would put all of my money, all of my effort behind TikTok and Instagram, which also happened to be, like I said, Instagram has not been living, seems to be disconnected from Facebook's operation here, and then TikTok's totally on their own, and they're starting from scratch. And so those are the places you'd be most effective, I think get me away with it. When you talk to people who are going at TikTok

who are, like you said, legit, are they concerned that it's owned by a Chinese company, that's interference there? I think they are well aware of issues there. I don't know of any examples where they've been overridden, but I wouldn't know either, right? So I think that is something they have to be concerned about.

Now, when it comes to the kind of this, like, disinformation around the election, I'm not sure that's relevant. I think where that's going to be more relevant is around whether or not they enforce their rules against Chinese actors. So, you know, if I would not see TikTok U.S. enforcing their disinformation rules against a Chinese state media. That's the kind of thing. But when it comes to like kind of the Russian trolling in the election, I think that's less relevant. A bigger issue there is that the Chinese companies have just not done as much investment in the trust and safety side. other than in political topics that are of interest to the Communist Party. And so I think that that's part of their issue. Yeah. All right, well, Alex, we have gone way over. I appreciate the extra time you've given us. You're at the Washington Post editorial. You can go find that.

You should read it. It's very good. Your colleagues have just written about Plandemic as well. Go check that out. And we'll be writing more about that too. Where else some people find you? They can find all of our reports at I.O. dot Stanford.

And I'm at Twitter.com slash Alex Demos. Always a pleasure. We'll have you back on. soon. Thanks, guys. All right, thanks to Alex Stamos for joining us. That dude is not shy. It is always so much fun to talk to him. Also, my thanks to Casey Newton, as always, we'll be hearing more from Casey as the weeks go on with the Vergecast here. We'll be back on Friday with a chat show, and then a really big interview coming up next week. I'm not going to give it away, but Dieter and I are prepping for it furiously. I'm really excited about it. That's coming up next week. You can tweet

at me. I'm at Reckless. Love hearing your feedback. We'll talk to you soon.