Defense One Radio, Ep. 77: Insurrection, social media, and the future of tech policy

Google Play Apple Podcasts

This episode, we try to better understand those hundreds of Americans who stormed the Capitol. We’ll also look into what “national security” means when an insurrection can be stoked so openly, as it was in the days leading up to Jan. 6.

Our guests are Robert Pape of the University of Chicago and director of the Chicago Project on Security and Threats (at the 2:45 mark); Divya Ramjee of the Center for Security, Innovation, and New Technology and Elsa Kania of the Center for a New American Security (at 11:54); and Peter W. Singer from the New America think tank (at 21:33).

Find a transcript of this episode below.

Subscribe either on Google PlayiTunes, or Overcast, or wherever you listen to podcasts. Thanks for listening!

Here’s something you may have missed. I know I certainly did over the past couple months. 

For the year that ended this past September, U.S. Attorneys' offices prosecuted 21 different cases on international terrorism. 

During that same 12-month time, they prosecuted 183 different cases in domestic terrorism. That was more than double the previous year. 

And those 183 domestic terrorism cases last year? That was the highest annual total since the Justice Department began tracking this stuff publicly — more than 25 years ago. (And if you wind back that clock, it’ll take you to right after the Oklahoma City bombing.) 

The top charge, the top domestic terrorism charge in 2020 — with 39 counts — was assaulting law enforcement. Again, that’s last year’s data. 

Since the so-called “failed insurrection” of Jan. 6, where five people were killed — including a police officer — a whole new slate of researchers have been digging into a whole new data set all about domestic unrest and the digital roots of insurrection in America.

And that’s what we’re going to get into today.

Because so far, the Justice Department has filed charges against more than 200 people variously involved in the January 6 attack on the U.S. Capitol. 174 men and 27 women, with the majority driving or flying from Texas, New York, Pennsylvania and Florida. And social media? It was used to help bring charges against more than 80 percent of the people arrested so far.

And those numbers could all grow by the time you hear this episode.

Today we’ll ask what kind of things we can do better from here. How we can, for example, better understand those hundreds of Americans who stormed the Capitol. We’ll also look into how now may really be the time to truly expand the way we think about certain key things in our wired world. Key things like what “national security” means when an insurrection can be stoked so openly, as it was in the days leading up to January 6.

“I've been doing work on political violence for about 30 years, sort of different countries around the world.” 

That’s Robert Pape. He’s a political-science professor at the University of Chicago and director of the Chicago Project on Security and Threats. 

“And starting last spring, I started to work on political violence here in the United States.” 

In retrospect, that’s incredible timing. So, why start in the spring? 

“Well, I became worried as the federal government dropped the ball on COVID. In April, I became worried about the loss of social trust.”

I called up the professor not for anything that happened last year. I called him up because his team of researchers have pored over what we know about insurrectionists who’ve been charged with crimes of some sort or another. And that happened just last month.

What his team found was fairly surprising in a number of ways.  

For example…

Watson: You write in The Atlantic, “The Capitol Rioters Aren’t Like Other Extremists.” And after looking over nearly 200 people arrested in that riot, you learned “89 percent of arrestees have no apparent affiliation with any known militant organization.” Were you surprised by that once the metric began to come into focus? It seems to me among the more remarkable findings from this work.

Pape: Yes, we were. So we were surprised by that finding, then, just like I'm sure a lot of people are, because up until the time that we had collected the research and what we're studying are all the court documents, the criminal complaints, affidavits statement of fact, compiled by the FBI and police to prosecute you these the these individuals here. So these are documents which are primed to be full of connections to militant organizations like the Proud Boys. And actually that study that came out in The Atlantic, we've now been updating here. So we're updating this on a daily basis, we're well over 220 of those arrested, and it's now 90%. So it's if anything, it's moving slightly in the direction further in the direction of the overwhelming majority are not affiliated with militant organizations. And this is why it's important to not jump to snap conclusions about who are the Capitol Hill insurrectionists. We have to remember that most people didn't think there would be a Capitol Hill insurrection. So it's very important to understand the demographics of who we're dealing with, if we're going to create truly viable solutions that are good for America.

Watson: Right. Well, to that same point, digging just another layer deeper, the employment factor is fascinating to me, too. I would have expected to find a greater number of unemployed arrestees; what you found was essentially the exact opposite. 

Pape: Yeah, so I that's right, Ben. So what we've found is that it's now 9% of those arrested for breaking into the Capitol building or the grounds are unemployed, that compares to 25% of right wing extremists who were arrested for deadly violence from 2015 to 20. In our country, and those who have been tracking political violence around the world for years here know that it's normal to find unemployment at the level of 25%, 33% in the pools of the demographics of the violent extremists. That's just simply not the case here. And what that does, is it’s really challenging a lot of our understandings of what's driving people toward the violence, and also solutions. So a big solution that's often bandied about for stopping political violence is, well, let's give them jobs. Well, if only 9% are unemployed, that's not gonna work.

Watson: Yeah, that is fascinating. Do you think then in this case traditional motivators — like economics or employment, they’re much different here — do you think that might be because of Trump’s national profile? 

Pape: Well, I think it's not just his national profile, but I think we have some strong evidence in the data that Trump himself has played a pivotal role in congealing the Capitol Hill insurrectionists. So we have statements by the insurrectionists themselves and their interviews to the FBI. That is there. They can be held accountable for that and also into them in the media that are in these court documents. There are dozens and dozens of the insurrectionists saying that they came to Washington, D.C., specifically because President Trump called for the loyal “patriots” to come to Washington, D.C., on January 6. They say they specifically marched and broke into the Capitol because President Trump called for them to do that on that day in the speech that he gave. So we see already strong, strong evidence that the overwhelming motive of the Capitol Hill insurrectionists is directly tied to the pivotal activity of then-President Trump.

Watson: What, 41 different states that were represented?

Pape: Yeah, so it's about 40 states here that are represented. And more troubling than that, it's 57% of the insurrectionists are coming from counties that [President Joe] Biden one. So they're not coming from the presumptive red parts of the country. Very few are coming from the places Trump won by 80 and 90%. Again, 57% are coming from counties that Biden won, and about half of those from large urban areas like Washington, D.C., Los Angeles, Dallas, Chicago — these are normally the blue, blue parts of the country. And what you're seeing is that in an area that these are pro-Trump, these are Trump supporters, in an area that Biden may carry 65%. Well, that's a slam dunk for a political victory. But that still means there's 35% who are supporting Trump. And in the case of say, Dallas, that's 300,000 voters for Trump and Dallas produced four insurrectionists. So what we're seeing here is not a picture that leads itself to thinking ‘Well, okay, if this was in a more mixed part of an environment, or an environment where the insurrectionists would be coming into contact with people of different attitudes, here, this, this would be another solution.’ Well, that's not gonna work here, either. So rounding up the normal, the right-wing militant groups, that's not going to be a core solution, because that's only a tiny percentage of the Capitol insurrectionists. Creating jobs is not going to be a solution here, because that's again, a tiny percentage. QAnon? It turns out only 8% of the Capitol insurrectionists are identifiable members of QAnon. So this is another, you know, thing that's in the news, which is not turning out to be systematically helpful. And so what you're seeing is over and over and over again, what you mainly see is we've learned a lot about what's in the movement here already, but we need to go further. And we need to thicken up our understanding of who's in this new mass movement, as opposed to trying to reduce it to the normal suspects that we always think is around political violence movements. Here, we need to resist going to pat solutions before we know more.

Watson: One of your report’s recommendations is for more widespread “de-escalation approaches for anger among large swaths of mainstream society.” What can you tell me about this course of action?

Pape: That's right. Well, keep in mind what we're mainly saying is obviously we want to think about the escalation of the anger; but the main thing, what we're saying is, we need to know more before we just shoot from the hip and think we're going to create solutions. So after 911, remember, a lot of people thought that we're going to stop terrorism by going invading and conquering Iraq. Well, that didn't work out too well. And part of the reason was because we decided on a solution before we knew much about the problem. This is a new problem, major political violence in the United States. This is a new mass movement that is congealed here, by the activities of President Trump. It's building on mass grievances here that are held by millions such as the election was stolen. There's now a pivotal focal point event of January 6 itself, which is being memorialized in the in the movement as the Independence Day or the sort of similar to the Boston Tea Party. Here too, as a rallying cry to mobilize more. Well, what this tells tells me more than anything else is this is not a movement. That's going to be over here in the next week or two, or a couple of months; this is likely going to be with us through the 2022, if not 2024 election cycle, and it's congealing. And what we need to do is we need to do more work to understand the movement here before we sort of start saying, ‘Oh, yeah, it's obvious we should do X, Y, or Z.’ I'm not saying we wait years, but I am saying that we need much more work. And that's what we had at CPOST. We've already been doing this. We actually have been doing this starting last since last April. And we're going to be doing more here over the next three to six months in a very, very energetic way. And that's what we need to do going forward.

Watson: Robert Pape is a political-science professor at the University of Chicago and director of the Chicago Project on Security and Threats. Professor Pape, thanks so much for talking to me.

Here are a few things we know about the riot and insurrection of January 6: It was coordinated across several different social media platforms. Politico reported on a significant portion of that just two days after the insurrection. Sites like Just Security have published work tracing the hashtag and planning trail much farther back, all of it out in the open.  

In terms of coordination and messaging, what was on Facebook in the run up to January 6, especially in its Groups sections, was alarming enough that Facebook has made significant changes in the weeks since. 

And then there is Parler. A new site; well, new-ish. Started in 2018. All about free speech. It’s attracted mostly conservatives, many of them in droves after Trump’s loss in the 2020 election became increasingly apparent

Parler has since become a gold mine for federal agents, especially after the site’s really poor security allowed one hacker to access all user videos, which were then released for public viewing (and not just for members of Parler) — assisting journalists and investigators looking into the events of January 6. 

Now those are several significant events that happened in just a few weeks time. And a lot of it — indeed most of it — played out on social media.

We’re not here to ask things like could Jan. 6 have turned out differently? It’s moot. 

But amid all this, I remain genuinely curious to know what’s next when it comes to the future of social media and national security.

America’s not the only ones thinking about it. French President Emmanuel Macron mentioned it when asked last week about xenophobia, violence and the state of liberal democracy today.

Macron: “My deep conviction is that social networks are definitely part of the roots of this change, which is once again an anthropological change, because they legitimated in a certain way the lack of inhibition in the different speech. They promoted, I mean, the culture of tough words, of conflict, and so on. And it progressively, according to me, changed the nature—the deep nature of what the democratic debate should be. This is why, if we want to preserve our democracies, we have to address these issues.”

I recently called up two U.S. researchers who’ve given the future of social media quite a bit of thought. Their article is entitled “Regulate Social-Media Companies.” 

Watson: Divya Ramjee of the Center for Security, Innovation, and New Technology — welcome to Defense One Radio. And Elsa Kania of the Center for a New American Security, welcome back. Divya, we’ll start with you. How do you describe the kinds of algorithms you write about to your family, or really just us non-specialists?

Ramjee: Well, you know, I'm glad you asked this question. Because after this article came out, my family actually did ask me, you know, what is it that the social media algorithms do? I think the easiest way to kind of describe it is just to simply say, algorithms in general, can use data to suggest an outcome based on that data. And so for social media companies, these algorithms use data from what you're viewing, what you're clicking, what you're scrolling through, and even some aggregate data based on people who are similar to you on their platform, to them suggesting content and keep you using the app.

Kania: And I'd add, my dad joined Twitter recently and was complaining to me about his timeline, and why was he seeing what he's seeing. So I had a similar conversation of explaining about the algorithmic timeline. And some of the ways the Twitter in particular often tries to promote tweets that are viewed as most relevant whether in ads or in the general content you're seeing as you're scrolling your way through the pandemic. So I think, when we try to capture in the pieces how, how trends and dynamics we're all familiar with, as users of social media platforms relate to some of these concerns of national security, that these aren't just creepy and annoying features of the platform's we spend far too much time on these days, these can actually be the patterns that reinforce radicalization or make that by design make these platforms more conducive to untrue and extreme content. 

Ramjee: We're very focused on the algorithms that social media companies use. specifically artificial intelligence algorithms, to keep consumers engaged with their apps, and not necessarily moderating the content that is being posted on the platforms.

Kania: Algorithms aren't magical, nor do they need to be so mysterious. And I think that there is an onus upon social media platforms — as they're trying to rebuild some trust and contribute to being part of the solution, rather than primarily regarded as one of the sources of the problem — certainly greater transparency would be very welcome on many fronts. And I think there's also a question of tech literacy and public education that all of us, as citizens, as consumers, as individuals, interacting on these platforms, whereby our data, and our identities sometimes become a de facto part of the product. I think we we all need to work towards having more of a baseline of understanding about how these, how these platforms work, the algorithms that drive them, and that can impact our own behavior as well. And also, I think there is also an onus upon policymakers to make sure that they have not just a baseline understanding, but the technical expertise and perspectives informing how they approach legislation and regulation. 

Watson: I like your idea of "impact assessments of algorithms before their launch, particularly for AI algorithms that learn from and leverage consumer data." How easy/involved might something like that be?

Ramjee: Well, you know, like everything, there's always bureaucracy; but it's really not something that we can't, in fact, implement. And there are countries who do actually have these requirements for impact assessments. I think we all kind of are familiar to some extent with GDPR, which is the EU’s General Data Protection Regulation. And GDPR essentially requires that there are data protection, impact assessments for technologies such as artificial intelligence algorithms that would pose a substantial risk of harm to consumers. And so when you already have other countries that are able to implement these types of impact assessments, you can kind of see that the U.S. is still quite behind in implementing these same kinds of standards for these technologies in our own country. 

Kania: Even as tech companies have signaled or hinted that they would welcome certain kinds of regulation and appear to be open to some of that, there's also the reality that in some cases, American technology companies are global behemoths that present some very complex questions for policymakers. And I think it's clear at this point that the trajectory we have been on is unsustainable, and a wholly hands-off approach to the internet, actually is not conducive to the freedom of speech we've been trying to uphold when there can be this amount of hate speech and just information and quite dangerous radicalization and polarization that really undermines the marketplace of ideas that we had once we had once perhaps naively hoped the internet would, would expand and enable.

Ramjee: I think, you know, also to build off of that, we're in this age of really intensified social justice movements. Within this, there have been increasing cries for improved privacy standards, and more transparency, from tech companies that are implementing these kinds of algorithms in our society. And even as you know, as mentioned, we've seen firsthand now, what these kinds of algorithms can help facilitate in the real world. And it came out in a very, very negative way, on January 6. 

Kania: I think the only way to remain private and the world we live in now and really ensure your privacy is to avoid having a cell phone in the first place and really eschew all modern technology and there are some people who are willing to go that far. But most of us for reasons of feasibility, do live in the world as it is and in which an impingement upon our privacy can be a feature and a quite persistent problem. And one of the problems beyond that, and I think reasons why there is there are compelling reasons for the U.S. government to really turn attention in earnest to data privacy and security is that the collection of that data can also be of benefit to any number of other actors, whether we're talking about cyber criminals or, or potential foreign adversaries. And if you care about privacy and civil liberties, and if you care about enhancing our resilience against threats in terms of counterintelligence, either way, questions of data privacy and security should be near the top of your agenda. So I think that at the end of the day, I think we do need to find ways to reconcile data privacy and security with national security, because our status quo is one of vulnerability and one in which the exploitation of data may cause unintended consequences down the road of for our national security and for our civil liberties and democracy on on any number of levels. 

Watson: Divya Ramjee of the Center for Security, Innovation, and New Technology, and Elsa Kania of the Center for a New American Security — thank you for speaking with me.

Peter W. Singer is a strategist from the New America think tank in Washington, D.C. Peter, welcome back to Defense One Radio. 

Singer: Great to be here. 

Watson: You’ve laid out what you call “Three Steps to Fight Online Disinformation and Extremism.” And this is after the pretty big moves like removing Trump from Twitter and yanking the Parler app from mobile stores like Apple and Amazon. But you write about “short-, medium-, and long-term efforts” to clean up our information spaces and hopefully drop the chances people become radicalized into Capitol-storming extremists. First, in the near-term, you write, we should change the message. Can you describe what you mean by that?   

Singer: So this is in response to how the narrative is turned in a certain way. So, you know, look, the companies have a great deal of problems. But now, while there are many valid concerns about the companies and their power in a great deal of the critique against them and the correct actions that they took, are actually bad faith efforts to shift the conversation to rewrite the narrative. It's about arguing that, you know, somehow, in the wake of six deaths, one of the most shameful episodes in all of American political history, that, ‘No, no, I'm the true victim of this, I'm the one being censored.’ And so, you know, this is about, Hey, you know, for the first time, we need to understand that what these companies did is right. And while we need to hold on to, you know, ensuring that they continue to implement their processes correctly, and the like, that let's not be distracted by this faux argument that's out there. I think more this, this discussion of speech is about understanding the differences between the mythology of certain forms of speech, free speech versus dangerous speech, as well as understanding what companies can and can't do in this space versus what a government can and can't do in a democracy. And most of this debate about censorship right now is a bad faith debate that is really about trying to reclaim the mantle of victimhood. I know that's hard for people to hear. And again, like I want to be clear here like there is a real serious discussion to have about everything from the role and power of these companies to what is appropriate and allowable speech online or not. But the vast majority of the discourse on it, after January 6 is basically people deliberately misinterpreting what censorship is, or isn't in an attempt to change the conversation and somehow make them the victim of the story. 

Watson: Your long-term advice struck a chord with me. I’ve been thinking about media, information and data literacy — and really the lack of robust programs related to it — for the past several years. You write, “Across America’s roughly 14,000 school systems, most schools lack digital literacy programs, while those that do have been largely on their own to find and pay for teaching tools that work.” Furthermore, as you have pointed out on Twitter: “In 85 proposals made by 51 different organizations exploring what needed to be done to battle against the online forces of mis/disinformation, the most frequently recommended policy action was to raise the digital literacy of those who consume that information.” You’ve been looking into this for some time, I gather. What have you discovered as far as why digital literacy and, as you put it, cyber citizenship programs — why are these so hard to find? Finland apparently has a sort of approximate-enough version of this. Now here in the states, if this had been a sticky politics of education question, I feel some of that stickiness is gone after Jan. 6. 

Singer: Yeah, so let's let's be clear about this. The problem of online toxicity, you know, whether it's misinformation deliberately spread disinformation, conspiracy theory, it's not just some partisan issue. Maybe there are certain elements of it that play out that way and you know, what you think about the capital takeover or whatnot. But look, that problem of misinformation hit every single other policy area. Whether it is why the pandemic has hit harder, it's because of the associated “infodemic” is what the experts call it of, you know, all the misinformation that surrounded Coronavirus. So it's a public health issue. There's a military side of it, it's been used to target not just our democracy and elections has been used to target individual military units when they deploy into NATO’s eastern borders. Oh, if you're a parent, it's used to target your kids to trick them into, you know, thinking a certain way to buying a certain product. So all of us need these skills to more effectively operate online to learn to discern as they say it between you know, not just what's real and fake but when am I being targeted or not? How might I be manipulated or not? What should I share or not? So you know, as the article puts it, it's not about telling people what to think it's not about fact checking; it's about giving people the skills of how to use this space effectively. And again, it's not about one single area, it's about, you know, changing the overall ecosystem. But it's also about you know, how we think about the solution sets. Basically, we're always looking to code as a solution, either software code, if only the companies would alter their algorithm to do X, Y, or Z, that would solve the problem, or we're looking for legal code, if only the government would do my preferred policy solution, if it would break up the companies that would solve the problem. The hard reality of it is first, given our politics, given the way businesses work, you're not going to get everything that you want, and those algorithmic or legal code changes? It's just not going to happen in our Congress; it's not going to happen in the private companies. But also, even if they got them, it forgets the human side of this — it forgets the human side of the attacker, ‘Okay, you change this, then I'm going to alter,’ and it forgets the human side of the target us. So that's one of the reasons we don't have it. 

Watson: Well that’s sobering. 

Singer: There's another part of it, I think, as Americans, not just in this space, but man across public health, military doctrine, you name it: We don't do a good job of learning lessons from other nations’ experiences. So, you know, we have to admit it. Look, there's other nations, you know, the Finlands, the Estonians of the world, that handle cyber threats better than we do. We ought to learn from them say how can we implement something certain aspects that they have? There's another part of the implementation problem is, which, you know, a Finland or in Estonia has an advantage against us is that they have national school systems in the U.S.; we don't have a single national school system. We've got a department of education that can support and coordinate; but you know, in the U.S., we have, by one count 14,000 different school systems — state, local county, some of them, you know, have hundreds of 1000s of students, some of them have hundreds of students. We also have an ecosystem that doesn't want to invest in the kind of the long-term even if it's effective. And then finally, it's difficult in our system, because it's just the way we run things in America. Each of those though, doesn't mean we can't have it. We can certainly fund it, we can certainly coordinate better. For people that know my work, they're like, ‘This is the guy that right, you know, wrote on robotics and war with China.’ But why am I so engaged in this topic, because it's so crucial to every single other threat that's out there, whether you care about foreign governments going after us and you know, hybrid [or[ gray-zone warfare, to you know, ‘I care about climate change,’ to ‘I care about terrorism and extremism,’ to ‘I'm a parent, and I care about my kids.’ This is one of those real action items we've got to go after. And so that's why I've been involved. And also the fact that it is an area where it's not just national security, it's not just education policy, it's not just tech policy — it's how they all come together. I don't think these issues have to be framed as partisan, or we can only act on them within four years. They weren't needed two years ago. They're needed now. There'll be needed 10 years from now. The question is, are we going to implement them now? So that we have them 10 years from now?

Watson: Peter W. Singer is a strategist from New America. Peter, thanks so much for talking to me. 

Singer: Thanks so much. Take care.