About the Guest:
Kate Starbird is a Professor at the Department of Human Centered Design & Engineering (HCDE) at the University of Washington (UW). Kate’s research sits within the fields of human-computer interaction (HCI) and computer supported cooperative work (CSCW). Extending from early work in crisis informatics, her research program has followed the phenomenon of online rumoring down the rabbit hole and into some of the toxic online spaces that are increasingly (re)shaping discourse, values, and politics around the world. Her work has revealed how online disinformation — i.e. the intentional manipulation of discourse for political gain — is inherently participatory, taking shape through collaborations between witting agents and unwitting (though willing) crowds. Dr. Starbird received her BS in Computer Science from Stanford (1997) and her PhD in Technology, Media and Society from the University of Colorado (2012). She is a co-founder of the UW Center for an Informed Public, which works to strengthen democratic discourse by building resilience to online manipulation.
Episode Transcript:
CORY BARKER: In April 2013, two bombs exploded near the finish line of the Boston Marathon. In the hours that followed. People across the internet tried to figure out what had happened. On Reddit, users began combing through photos from the scene. They zoomed in on faces, circled backpacks, compared timestamps, built theories. It wasn't long before one name started to circulate: Sunil Tripathi. He was a missing college student. Someone found his photo online. Others began connecting dots. The theory spread quickly across Reddit, then onto Twitter now X, but it didn't stay there. News organizations picked it up and his name was broadcast as a possible suspect. But there was a problem. He had nothing to do with the bombing. By the time authorities identified the real suspects, the damage had already been done. A grieving family was pulled into a global news cycle, build on a false claim that thousands of people had helped construct. No single person made that mistake. It emerged from the crowd, from people trying to make sense of the chaos, trying to find answers, trying to help. But in the process, they created a story that wasn't true.
MATT JORDAN: That's what researchers now describe as participatory disinformation. And it's the focus of the work of Kate Starbird. She's a professor at the University of Washington and co-founder of the Center for an Informed Public, where her research examines how online crowds, platforms and information systems interact during breaking news events, crises and political moments. Kate's work shows that misinformation today isn't just created and then spread. It's shaped in real time through participation. People interpret, remix, challenge and amplify information as it moves. Often blurring the line between audience and actor. We're going to talk with her about participatory disinformation, what she's observed on platforms like X, and how these dynamics influence what the public comes to know and believe. Kate Starbird, welcome the News Over Noise.
KATE STARBIRD: Thanks for having me on.
MATT JORDAN: Sure. So, I want just to get a sense of how you came to study misinformation. You're a computer scientist. What drew you to this field of study?
KATE STARBIRD: Yeah, actually started out as a PhD student studying crisis informatics around 2008, 2009. And we were actually just looking at the use of social media during crisis events. And my dissertation was actually about all of the pro-social things people do during crisis events. I wrote a dissertation on digital volunteerism, how people would use Twitter and other tools to try to help other people during disaster events. So initially, I was again focused on all the good things that people were doing online. And then around 2012, 2013, we were still looking at crisis events. I kind of moved on from a PhD student to my own position as an assistant professor, and we were looking at, I think, data from the Boston Marathon bombing and began to see that rumors and misinformation were a bigger and bigger part of the convergence online after these crisis events, or even while the crisis events were going on. And so, from 2012, 2013 to about 2016, I was looking and focused on online rumors. And we don't even use the word misinformation as much as rumors. We were really kind of looking at that as just a natural response to crisis events is to try to figure out what's going on, and people sometimes get that wrong. So, we weren't necessarily putting a negative or normative framing on that kind of information sharing. But around 2016, 2017, we began to realize we weren't just seeing accidental rumors and misinformation, but we were really looking at pervasive disinformation, intentional exploitation of these spaces that was really sinking into the networks and the algorithms in these online spaces. So, it went from rumors and kind of talking about things to misinformation to really focusing on disinformation after about 2017. Now more and more looking at propaganda as an umbrella term, because so much of this is hard to just put in a disinformation framing. There's so much, especially with the onset of AI.
MATT JORDAN: And what is it about crisis situation? What insight does it give us about how online environments work?
KATE STARBIRD: Rumoring is actually a natural response to crisis events, in part because of the uncertainty, the crisis event, whether it's a manmade crisis, a terrorism or something, or even natural disasters, the early moments of the aftermath of those events, or when those events are ongoing, you might have a lot of information, but you don't know what's true yet. There's a fog of war situation. There's just a lot of uncertainty. And often there's also high anxiety. These are high stakes situations. Getting the right information can be really important for making decisions and so people have a tendency to come together in the aftermath of crisis events, to try to figure out what's going on. And historically, that was a come together in person. Now that's a come together in online spaces. And that process, we talk about it as collective sense making of like making sense of what's going on. You don't have perfect information. You're trying to put it together. You're making theories, you're speculating about what's going on. And then that leads to rumors. And rumors can sometimes turn out to be true. Often, they turn out to be false or somewhere in between. But rumoring is a natural response to crisis events. And so that's an interesting time to study that kind of behavior, because you've get these bursts of activity. And so, as a researcher, you have these like really short windows into this sense making activity. And again, like initially we entered the space not necessarily saying this is something that's wrong. I mean rumoring serves informational purposes that serves psychological purposes and community building. There's a lot of reasons people would come together and try to figure out what's going on. Additionally, these crisis events became opportunities for people to exploit those situations. We've always known in crisis events, there's a bunch of different types of people that converge, and one of them is an exploitation kind of persona, and these became windows of opportunity online for people to gain attention, to put out clickbait content, breaking news, and also political point scoring kind of narratives, to grow a reputation, to get more followers and then in some cases, to use those followers later for different kinds of whether it was financial gain, political gain or something else. So, on top of it being a natural time where people are converging, they're coming together. Lots of attention, lots of anxiety, uncertainty. There was also this kind of exploitation that we've seen. This new class of influencers take advantage. Many of them take advantage of these crisis events to grow their audiences.
CORY BARKER: Speaking of coming together and exploitation, one of the great concepts from your work that you explore is this idea of participatory disinformation campaigns. Can you describe that for us and for our listeners and a little bit more detail?
KATE STARBIRD: Yeah, I mean, I think if we go back in time to maybe 2016 when the word disinformation all of a sudden, if you look at the graph like nobody's really using it, even we were studying rumors and misinformation, and we sometimes use the term disinformation, but rarely didn't quite really know what it was. It really takes off around 2016, with in particular, the Russian government's intentional campaigns to manipulate information spaces. And then other folks realized they could use this Russian style of disinformation and that they could manipulate information spaces in this way. And so, it becomes ubiquitous, not just the term, but the use of the techniques become much broader than just one set of actors. But in 2016, when we were looking at it, we were looking at through this lens of the Russian state actors and what they were doing, these deceptive campaigns. And there was a lot of talk about this really top-down mechanism where these, these knowing actors would manipulate these information spaces. And as we continue to look at things, yeah, we sometimes could infer that there was this top down. There were bots or deceptive accounts that were organized in some in the Internet Research Agency, like an actual building in Russia, that these folks were being trained there and in manipulating the spaces. But by the time we get to 2020, we start that even before that, our work is like, this isn't just top down, it's also horizontal. It's also bottom-up. Like everyday people are becoming unwitting agents in the spread of this content, and then they begin to create it themselves. And it's really hard to disentangle the witting actors, the people that are in on the campaign, with the people that are just becoming participants. And so the first time I started writing that, it was 2017, 2018, and our first papers in 2019, where we're really starting to look at disinformation as a participatory process, as both top down and bottom up, where online audiences are collaborating with influencers and with the perpetrators of disinformation campaigns to create content that broadly fits into something that can be misleading. More and more, I'm thinking about participatory propaganda because it's hard to fit it quite into disinformation. Some of these campaigns, but the idea is that they're not just top down, but they're these collaborations with different sets of actors that have different motivations, that are participating in different ways. And many of them may not know that they're part of a disinformation campaign, especially to the audiences who begin to sincerely believe this content. And even when they're not sincere believers, they're unwitting but willing. They want to believe this content, and they begin to become participants in in these campaigns. And we can give examples of the 2020 election denialism or claims about voter fraud, and we can unwind them and see that they're untrue. But a lot of people that were even generating them really believed that they were being cheated. In some of these cases, more recent cases that they're eating the pets claims and in 2024 around immigrants. And we can see people, some people believe it and they're spreading it, and then others are just they're sharing this content. They don't necessarily believe it, but it fits into their either their political goals or their worldviews.
MATT JORDAN: I've heard you describe this before, especially in relation to the right-wing media ecosystem as operating like improv theater. What do you mean by that?
KATE STARBIRD: That's it's interesting. So, a couple of things there that kind of set up that argument. One of them is there's this reflection at the 2024 election. The Democrats lose badly. And there's a lot of talk about how the right dominated certain sections of media and certain demographics that were using different kinds of online media, podcasts, social media. And we are actually looking at that moment, but we also are sitting in our data from 2020, and we're trying to think of like, what is the right doing differently in these online spaces than the left? And one of the things we see is that the left is still they're mad at The New York Times, a lot of them, for not reporting on things properly according to their ideas of how things should be reported on, but they're still heavily reliant on traditional mainstream media outlets. And the right has really embraced a very different kind of content production with online influencers. A lot of hyper-partisan media outlets that feed into this ecosystem, and we begin to see and it's across a bunch of different research that's happening. So, I have I'm working with a PhD student, Anna Beers, and Anna was looking at these networks of influencers and how they often had shared audiences. So, an influencer wasn't just surrounded by if you look like the network structure, an influencer isn't just surrounded by people that follow them. And there's another influencer with different people, but it followed them that their audiences were shared in some interesting ways, and they often were interacting with each other, and they were playing different roles in the information ecosystem. And this is primarily on the right. The left has this too is just not as well developed. And so, we see like that as part of the more participatory model like the right has done more participatory politics and I can see it disinformation. They're going to see it as political messaging. We can have a normative argument about what the value system is represented there, but they're doing it better in these online environments. But part of that is it's not just a top-down message. There is some top-down messaging, but there's also this empowering of audiences to be part of the show. And by being part of the show, we've got these influencers that are on stage with these shared audiences. But the influencers don't just send their message out to the audiences. The audiences actually shape what the influencers are doing by giving them feedback, by seeding stories, by saying, hey, you know what? I saw somebody in my neighborhood, and I think they were eating pets. Or in the case that we were seeing, we had. So, a lot of people, like the Sharpies were bleeding through my ballots. Is that voter fraud? And then the influencers are like, yes, that must be voter fraud. And picking it up. So, we see that the audiences are able to contribute to the show in a way that aligns with the participatory nature of online environments. It's very effective. It empowers audiences. Audiences feel like they have agency in the political messaging. And this in 2024, when we started conceptualizing this and we had a few papers earlier, we start to talk about improvisation on the right and how their messaging is more improvisational, which matches to the like dynamics and logics of these online environments. And so, then we kind of say, oh, it's kind of like improv theater. So, we have a metaphor, Danielle Thompson and I kind of fleshed that out a little bit for paper, but I feel like it's an interesting way to think about how different political parties have leveraged these dynamics differently. And maybe unintentionally. It's just part of the way these parties have evolved, in part because the right had to, in their minds, use their use different media ecosystems because they didn't feel like mainstream media were serving their needs. And the left stuck to this mainstream media, which probably were not no longer serving their needs, but have had a hard time figuring out how to develop a similar kind of ecosystem. There's a danger in participatory politics where they can spin out of control, where audiences can radicalize their influencers.
CORY BARKER: If you're just joining us, this is news over noise. I'm Cory Barker.
MATT JORDAN: And I'm Matt Jordan.
CORY BARKER: We're talking with Kate Starbird from the University of Washington about participatory disinformation, how it spreads, and what it means for the information environment. I'm curious, given your great work underlining the role of President Trump in the way that you see his role evolving as this ecosystem is evolving? I'm thinking so much about our current predicament with the US war on Iran, the ways in which every day he says something to the effect of like, the war is over, we've won. And then a few hours later, there's some post or public comment that actually we're going to escalate and we're going to double it right in the way that that is straight up lying or confusion or this yo-yoing effect that it's hard to some of your earlier points to put into a box of like, disinformation. So how do you view President Trump in this ecosystem in 2026?
KATE STARBIRD: Yeah, in 2026. This is I think it's changing. And so that's I'm going to start with how you've ended that question. I mean I can look back and we can look and see like President Trump, pre-President Trump, so, Donald Trump prior to becoming president the first time even in between like he's played different roles in this ecosystem over the course of his life. And over the course of his political career, I would say initially his style of political communication was just really well matched for our dynamics are like online dynamics, and part of that is his improvisational style is ability to kind of talk and kind of bounce things around and not be really necessarily committed to sticking by the thing that he said before, which we're now seeing pretty acutely. There's a different term that I haven't mentioned yet. We've talked a little about disinformation, mentioned rumors, but there's also the term bullsh*t. And there's a there's a whole book on bullsh*t, which I find to be very revealing. I can't remember the author right now, but like, it's I think it's a really valuable term. And I'm not using to use a profanity. I think there's a definition here is really important. And that is like content that you're saying, like claims that you're making, you don't care whether or not they're true or not. Bullsh*tting is about saying what is useful to you, whether it's politically useful or personal use for whatever, whichever, but bullsh*t is content that the truth value just isn't important. It's all about the persuasive value or the rhetorical value or something else. And so, Donald Trump is a very good bullsh*tter. Like, it doesn't matter to him whether or not it's true that he's trying out different things, trying to see what resonates with his audience, which is one of the reasons he's so well adapted to this online environment where you can get that feedback and he's happy to shift to something else. I can't speak to his mental and intellectual state here in 2026. I think there is something very different about his position right now. He's not as central inside the networks, but he has his own platform and his content goes out everywhere, and it both sets the stage to a bunch of people who are going to try to, like, improvise with it, either apologize for it, explain it, or celebrate it, whatever it is. So there's still interaction between what he's doing and all these folks. But also at this point, there are so many other voices, whether within the government or like this is the there's a thin line between the influencers on the right and the people that are actually in our government right now, which is one of the fascinating things that from someone who's been studying this rhetoric and some of these folks for so long, to see them in positions of power, but they now have accounts that represent other parts of our government that are doing similar kinds of improvisational and propaganda messaging, like Donald Trump's account is now complemented by these others in a way that, yeah, I don't have a simple way to characterize it. It's one of those things that probably won't become clear to folks like me for a couple of years. So, I'm not going to be super helpful here. But certainly, like there's a dynamic that's changing, but there's a recognition that like, one of the reasons that Donald Trump has been president of the United States twice now and in this era is that he's been particularly good at the kind of political rhetoric or bullsh*t that is just really effective in these online spaces. And he's been supported by people who have made sure that his message and his messaging style is that he doesn't pay a political price for the negative impacts of that kind of messaging.
MATT JORDAN: I heard a story on NPR the other day about how a lot of people around the globe were adopting his communications strategy. In particular, this was about how Iran had adopted trolling as a propaganda strategy. Why is trolling so effective, given the affordances of our digital platforms and media ecosystem?
KATE STARBIRD: The hard thing is like trolling means different things in different areas. There's trolling initially meant just trying to get a rise out of people, trying to get them to be upset about something. And that's been a feature of online environments. One of the first studies of email they already were seeing where people would get mad at each other and say negative things. And that was not even anonymous. And they theorized that it had something to do with because you're not in person, you can't see necessarily the results, like some of our natural empathy or our natural moderation of our speech goes away in these online environments. So even like just the like the agitation and the ability to say things that are harmful is different in online environments. But then you have the layers of anonymity. So, the anonymity likely contributes to people being able to say things without consequences. And so, people can both experiment but also violate prior norms and then also use it for different kinds of exploitation. More recent definitions of trolling often include the same anonymity driven behaviors. But then there's just like the changing of norms, these online environments have developed, the infrastructure of them is developed at the same time as the norms around what kinds of social interactions are okay or healthy in different ways.
MATT JORDAN: There was this moment after January 6th where it looked like there was going to be this attempt by internet platforms to clean things up. They kicked off some of the trolls and some of the worst spreaders of insurrection type rhetoric. They put a lot of money into their trust and safety teams. And it looked like for a moment, what are things where things are going to get better? What happened? Why did they just say, “Ah, we were just kidding”?
KATE STARBIRD: I mean, I think this is the stuff of many, many dissertations in the future. I would say that Republican strategists and communicators were very effective after January 6th. They needed to switch that narrative. They wanted to change who the heroes were and who the villains were. And so, they redefined what happened on January 6th by saying that the real perpetrators, a real bad thing, was not this crowd of people that were motivated by falsehoods to go and try to disrupt the results of election. The real problem was the internet companies and the people and organizations that had tried to stop those lies, or they tried to counter those lies and tried to call them out. And so, they managed to convince themselves and actually a large number of other people, that the problem wasn't January 6th. The problem wasn't the lies that Donald Trump and others told to motivate and mobilize the mob. On January 6th, the real problem was censorship, that somehow the truth was censored about what it actually happened in that election, and that that was the real issue. And they extended this, this kind of blended into some of the things that happened around Covid. And it turned out to be really, really effective. They tried other ones. They tried. No, it was really antifa. They tried a lot of other things, but they really did a great job of shifting the narrative. Around January 6th from these falsehoods were bad, and this was a bad thing that happened. And look at the results of them. And then, oh, no, that the bad thing was actually that anyone ever tried to do anything about that. And it really effective. And that that narrative goes on today. Right? Donald Trump and the current U.S. government just settled a lawsuit with themselves, with their own plaintiffs around so-called censorship by the Biden administration, which is a farce. If you look at the actual settlement, it tells them that they're not allowed to do things that nobody ever did. But anyhow, it's just really interesting. They've used they use the courts; they used social media outrage and different kinds of things. And they used to be honest; I've been studying this for a long time. People always thought they were being censored in, in, in online environments. It was a folk theory by many people. Back in 2010, after the Haiti earthquake, I was seeing these digital volunteers and they were like, oh my gosh, my content isn't getting out there and being censored for my political views. And it turned out that they were posting so often that the filters of Twitter thought that they were spam, and that they were labeling them spam and taking them down. So whether it's been true or not, people for all time had these really strong folk theories and I think the Republicans in Donald Trump and his supporters were very good at leveraging that folk theory and turning it into this deep story of censorship where their audiences believed it, their influencers were happy to propagate it, and they had a legal apparatus that went about trying to make that narrative into something that that a large number of people believed and it's their fight against censorship is still going on, even though even as this administration censors and attacks freedom of speech like no other in the history of the United States. So, it's been an interesting evolution to watch, both from the outside and unfortunately, as a researcher in this space, from the inside as well.
CORY BARKER: In a recent Substack post about Trump's claims related to the Iranian interference into the 2020 election, you have an aside where you explained that Iran's efforts were an influence operation, not an interference one. How should we be thinking about those as two different things?
KATE STARBIRD: Yeah. So, this is in reference to 2020. The 2020 election. Our team was and we put it go back there and I'm going to give you a little context. Our team at the University of Washington was collaborating with team at Stanford, and the team at Stanford had researchers who they identified that there was an Iranian influence operation in 2020, where agents of the Iranian government impersonated Proud Boys. So, they pretended to be Proud Boys, and they wrote intimidating letters and sent them to Democratic voters. So initially, we have all these voters saying, oh my gosh, the Proud Boys are harassing us. And then these, these groups at Stanford and other places identified that this is a Iranians are pretending to be Proud Boys. This was exposed. It was exposed very quickly. It was, and the organization at the time inside the Department of Homeland Security, there's an office called CISA, which is Cyber and Cybersecurity and Infrastructure Security Association, something like that. And they put out bulletins. They made sure people knew, they made sure people understood it. So, this was almost a nothing burger, right? So, like one in the end, it didn't actually intimidate people.
And it's hard to know what the Iran government's motivations were in terms of who they were trying to help with that it may have just been just trying to make people less trustful of the process. But you fast forward to Donald Trump statements as the US is doing this invasion of Iran and bombing of Iran or whatever this operation is, Donald Trump tries to use this claim of Iranian interference in the 2020 election as part of the motivation for this attack in Iran. And it doesn't make any sense. But it ties back to the repeated claims by Donald Trump, which he continues to make to this day that the 2020 election was rigged and that somehow, he's trying to tie in the Iranian government to being part of that rig, which is asinine. It doesn't make sense. And at the same time, there's this huge hypocrisy at the core of that, because the people that helped point out that campaign and make sure it was identified and stopped got called censors later by the Trump administration and silenced, defunded. And the Stanford Internet Research Agency is no longer; the group that actually helped call out that campaign. So, the layers of hypocrisy around his attempt to use the influence operation from Iran as part of his motivation for the war is really frustrating.
MATT JORDAN: It's interesting in America, especially after World War I, developed a very blossoming propaganda studies system. All these influential intellectuals like Walter Lippmann and whatnot were part of these initial studies. And these things were rebuilt during the World War II and stopped about 1942. But you've started to introduce that word propaganda again into the discourse. And a lot of times people stay away from that word. And like you said a second ago, you'd think that this would be a time where that would be something we'd really want to talk about. With the ability of Iran and China and Russia to get their information over the borders into our information ecosystem, you'd think this would be something we'd want to be doing. And yet, any time do you do want to do that, or you do that or anybody else does this. It's called censorship. It’s just a very strange moment that we're living in.
KATE STARBIRD: There's a lot of experimentation happening around propaganda, disinformation, manipulation, these online spaces, the folks that are doing that experimenting are getting very good. And they probably know a lot about how it works. They're able to get pretty quick feedback about how effective they are. Foreign governments, our own government influencers, any number of people are very good at doing this. To study it and to call it out and to draw attention to how it works. For some reason, that is the thing we're not allowed to do right now, though, I have students banging on my door wanting to get in and study these kinds of things undergraduates, PhD students. So there a lot of a lot of people recognize it's really it's really a need we have as a society of understanding what's happening is information systems helping people navigate them, developing new information literacies, developing technological solutions that layer on to the platforms we have to help people see, like where information comes from, what's the information provenance? Who created this? How did it go viral? How do I do this? Add AI it’s just head-exploding. There's so much need. And at the same time, we have a government who has basically tried to defund every group doing this kind of work and not just defund us, but like smear us, pull us into to interrogations in Washington DC. I've personally been through two of those, and it feels purposeful. I mean, there's also the chance that they just may be that they believe the conspiracy theories that that we somehow were pulling strings to censor millions and millions of tweets, which is just absolutely untrue. It's hard to understand, except as like the people that are gaining power through the use of these techniques propaganda, disinformation, manipulation of online systems, don’t want the rest of us to understand what's happening and to think about what we could do about it. And that's where we're at.
MATT JORDAN: Kate Starbird, thank you so much for joining us and for sharing your wisdom with our audience.
KATE STARBIRD: Thank you. This was really, really fun.
MATT JORDAN: That's it for this episode of News Over Noise. Our guest was Kate Starbird, professor at the University of Washington and co-founder of the Center for an Informed Public to learn more, visit news-over-noise-dot-org. I'm Matt. Jordan.
CORY BARKER: And I'm Cory. Barker.
MATT JORDAN: Until next time, stay well and well informed. News Over Noise is produced by the Penn State Donald P. Bellisario College of Communications and WPSU. This program has been funded by the office of the Executive Vice President and Provost of Penn State and is part of the Penn State News Literacy Initiative.
[END OF TRANSCRIPT]
Episode Credits:
Producer: Lindsey Whissel Fenton
Audio Engineers: Mickey Klein, Scott Gros, Clint Yoder
News Over Noise is a co-production of WPSU and Penn State’s Bellisario College of Communications. This program has been funded by the office of the Executive Vice President and Provost at Penn State and is part of the Penn State News Literacy Initiative.
