Invisible Rulers: Information Warfare and Public Trust

In her book, Invisible Rulers: The People Who Turn Lies into Reality, Renée DiResta shows the new ways propaganda is created and circulated to the public. It flows through influencers, algorithms, and networked online communities that can spread and magnify misinformation at scale. Different groups are often targeted with tailored messages, which leaves audiences with entirely separate versions of reality, making it harder to agree on basic facts or know what’s true.

DiResta discussed the challenges facing society at an event sponsored by the Corporations and Society Initiative (CASI) at Stanford Graduate School of Business. Moderated by CASI student leader Dylan Ackerman (MBA ’26), who previously spent five years at Facebook working on misinformation and voter interference, the conversation examined how “misinformation at scale” is eroding trust in institutions and straining democratic life.

DiResta explained that she became aware of the problem in 2013 as she was studying vaccination rates for preschools in California. This led to her to track public health data alongside the growing anti-vaccine movement online. What she found was quite concerning —social platforms were not merely hosting conversations, they were actively shaping them with automated accounts that created the illusion of widespread agreement. As once-fringe anti-vaccine views began to take hold, DiResta saw that public health officials were largely absent from these online spaces, which allowed misleading narratives to spread unchecked.

Her early work soon expanded to a broader examination of information warfare. DiResta was asked to advise the U.S. State Department on the online rise of ISIS and later helped inform the Senate Intelligence Committee’s investigation into Russian interference in the 2016 election, analyzing how state actors used the platforms to shape public perceptions at scale. In 2019, she joined Stanford’s Internet Observatory as Research Manager, where the team studied online trust and safety, investigating how information flows across platforms and how manipulation campaigns operate within digital ecosystems. After five years, the group was dismantled following intense political and legal pressure over its work on election integrity and COVID-19 misinformation.

As Ackerman noted, DiResta had been working in this space for more than a decade before “misinformation” became a formal field of study. In those early years, the work was informal and collaborative, with a small group of data scientists comparing observations and trying to make sense of emerging patterns across platforms.

To explain the complexity, DiResta shared a framework for understanding how misinformation operates in practice, developed by fellow researcher, Camille François. This approach breaks the problem into three interacting parts: actors, behaviors, and content. The first asks who is behind a message.

“Sometimes it's quite obvious, an influencer makes a piece of content, and you know exactly who they are. They're real, they're human, they're quite transparently who you are choosing to engage with,” she said.

“Sometimes, though, you'll see accounts pretending to be something they're not.”

That distinction, she noted, is critical in identifying inauthentic actors, including coordinated networks that impersonate communities or identities to gain credibility and influence.

The second dimension involving behaviors focuses on methods that determine how content is amplified. This can be achieved through bots or other forms of fake engagement and through coordinated networks that are designed to create the illusion of popularity or consensus.

The third factor, content, is more complicated, as platform rules about what’s allowed are often unclear and frequently shifting.

“You can't tell people to vote on Wednesday, and not on Tuesday. You can't tell them that their polling place closed when it's not, that's voter suppression. You can't delegitimize an election, so they do have some policies that they put out that are specific to certain types of content and certain types of areas.”

Following the 2016 election, as concerns about Russian interference arose, DiResta described a change in how researchers, government, and the platforms worked together. What began as skepticism hardened into something more openly adversarial. Independent researchers and data scientists were pushing platforms to recognize the scale of manipulation, often pointing to evidence, like coordinated networks driving widespread engagement. This evidence contradicted what the companies were saying publicly.

This tension played out in calls for congressional hearings and public accountability, as outside experts pushed both companies and policymakers to confront what had happened and explain it transparently.

“For me, a lot of it was sitting with Senator [Mark] Warner and others and saying, you've got to have hearings. Make them account for what happened.”

But that stance, she noted, had limits. While investigations conducted solely from the outside could identify suspicious patterns, researchers like DiResta needed access to the internal data to confirm who was behind it and fully understand how those campaigns operated. Over time, that realization led to a more collaborative model: platforms built internal investigation teams and, in select cases, shared data with vetted researchers under strict conditions.

“That meant we had the capability to go and do much more longitudinal research or trace things back to other forms of media, and the platforms would do the deep investigations into what happened on their platform.”

Importantly, the relationship retained enough independence for researchers to challenge platform conclusions and publish their own reports. For DiResta, this balance offered a more credible and accountable model for understanding and exposing coordinated manipulation online.

DiResta argued that efforts to curb misinformation by simply removing content often miss the mark and can backfire. Drawing on her early work tracking anti-vaccine communities, she noted that takedowns can reinforce a sense of “forbidden knowledge,” making content more appealing and deepening distrust.

She supports removing clearly inauthentic actors, such as coordinated networks posing as local voices, but is more skeptical of broad content moderation, especially without transparency. In her view, platforms should focus on disrupting manipulation tactics rather than deciding what counts as acceptable speech, which is often fraught with problems.

“There would be takedowns where the person didn't have a right to appeal. They don't really know why their stuff has come down.”

DiResta argued that platforms are not neutral conduits but active curators. Every piece of content is ranked; algorithms then determine what users see and in what order via recommendation engines. She noted that, under U.S. law, platforms have broad discretion to decide how content is ranked and presented.

“The platform has, completely at its own discretion, the right to decide how it is going to rank. That in itself is an incredible power,” she said. “Users should have much more control over that ability to do that ranking. But on a centralized platform, the right to curate is the platform's First Amendment right. That is what you were signing up for.”

Because ranking systems are built to drive engagement, they often end up amplifying division, pulling people deeper into echo chambers and reinforcing polarization in the U.S. and beyond. Ackerman picked up on that point, noting how easily platforms can lock users into self-reinforcing information loops.

DiResta pointed to a coordinated effort beginning in 2018 to frame content moderation, even light-touch measures like labeling, as censorship, pushing influencers and their followers to alternative platforms.

“The question of how do you bring people back together? How do you depolarize them? That's bigger than tech.”

She highlighted the concept of “bridging-based” recommendation systems that tend to promote content that appeals across ideological lines over content that amplifies outrage or sensationalism.

“You're still surfacing the issue, you're still getting the content out there, but you're not rewarding the rage entrepreneurs by putting them at the top of the feed every time.”

DiResta described the past few years as a turning point shaped less by changes in technology than by a sustained political backlash against the institutions studying it. In 2022, as control of the House shifted to a Republican majority, a narrative began to take hold that researchers, platforms, and government agencies had coordinated to censor speech during the 2020 election.

“The election stuff became very, very polarizing. And that was because of the 2020 election and the false allegations that it was rigged and stolen. That was where you started to see the politicization of that,” she said.

DiResta traced that claim back to a single, unsupported source that was amplified across media channels, creating an echo chamber of repetition that gave the appearance of independent verification. By early 2023, that narrative had translated into formal action: subpoenas demanding years of communications and investigations framed around allegations of a so-called “censorship regime,” despite the lack of evidence and the bipartisan nature of earlier work. The consequences, she said, were immediate and far-reaching.

“In my opinion, that was the entire point of the endeavor. It was election deniers mad about what we did during the election, pushing apart the entities that had collaborated to try to understand and triage rumors in the election.”

“And that was the goal. It was a very, very effective thing because I think all of the different entities retreated to their respective corners and said, it's too much of a liability to even speak to each other anymore.”

Legal pressure, subpoenas, and lawsuits created a powerful chilling effect across researchers, universities, and platforms. What had once been a coordinated effort to study and limit these risks began to splinter, weakening the networks working to bring transparency to the information environment.

Ackerman closed by asking whether the legal and political backlash had meaningfully changed how platforms approach moderation, or whether it simply confirmed a more cynical view that they were never deeply committed to these efforts in the first place.

DiResta pointed out that the platforms have been sued, too, but they have largely prevailed in courts, with judges consistently affirming their right to moderate as a form of editorial judgment. Despite the legal clarity, DiResta believes they should be doing more to win back public confidence.

“They're winning in the court cases. But they are not willing to stand up and say we made this choice, here is why we did it, this was the ethics underlying it.”

In DiResta’s view, the silence is due to sustained political pressure from multiple directions, first from the left during COVID, then from the right in the pandemic’s aftermath and claims of a stolen 2020 election.

As the conversation concluded, DiResta returned to the core idea behind Invisible Rulers: power in the digital age is no longer held by institutions alone. Networks and digital systems play a major role in shaping what we see and believe. With few reliable tools available, people must often depend on online communities, personal circles and their own judgment to decide what sources of information can be trusted in an increasingly uncertain world.