Just How Influential Are Foreign Governments’ Social Influence Campaigns?

These days, it’s all too common to see a front-page story about a foreign government’s influence operation — secret attempts to sway the opinions of another country’s citizens through social media campaigns, paid advertising, hacking, direct emails, or SMS text messaging.

In August, the FBI confirmed that the Iranian government was behind a hacking scheme to breach and subsequently leak confidential information about both the Trump and Harris presidential campaigns. Last week, the FBI reported that the operation is likely ongoing.

These attempts to infiltrate and influence the American public are nothing new, though.

In fact, Jacob Shapiro, a professor of politics and international affairs who co-authored version 5 of the "Online Political Efforts Influence Dataset" through Princeton's Empirical Studies of Conflict Project, said that his team cataloged 107 similar attempts by foreign governments since 2011 alone, with Russia, China, and Iran the lead perpetrators.

Now, a three-year grant awarded from the Minerva Research Initiative at the U.S. Department of Defense and United Kingdom Research and Innovation is giving Shapiro and colleagues at Columbia University and the University of Cardiff the opportunity to collaborate. Together, they’ll advance understanding of how political influence operations have evolved in recent years, the state of current operations, and their measurable impact on people’s behavior.

"An influence operation is fundamentally a way of trying to shape how another government or its people think about a set of political activities. The problem we are trying to solve is to determine if these campaigns threaten democracy or not. Are they actually moving the needle in our politics?" —Jacob Shapiro, professor of politics and international affairs

Dubbed the Influence, Manipulation, and Information Threats as Adversarial Techniques: Events, Evolution, and Effects (IMITATE3) program, Shapiro is partnering with Martin Innes, director of Cardiff's Crime and Security Research Institute, and Tamar Mitts, assistant professor of international and public affairs at Columbia University.

IMITATE3 will work to measure the impact of mainstream media and social media-based influence models and methods in real time, focusing on several countries with elections scheduled for 2024 and 2025, as well as ongoing conflicts, including the Israel-Hamas war and the Russia-Ukraine war.

Mitts said they chose to focus on contemporary information manipulation campaigns for a practical reason: "It gives us real-time data collection and allows us to capture information that is posted online before it's taken away. In many cases, when we are talking about content considered harmful or in violation of platform policies, it often ends up being taken down, and we can't study it anymore."

IMITATE3 is broken down into three task areas: First, the team will use open APIs and other entry points to gather systemic, large-scale data from social media platforms, and other sources to see how governments involved in these conflicts are activating information campaigns.

Next, they'll deep-dive into 2024 elections worldwide, examining how information campaigns were deployed — or not — in more than 100 countries.

Finally, the team will develop a standardized framework for comparing how influence operations work and measuring different countries' efforts and methodologies using current conflicts as their case studies.

Previous research by other teams largely involved quantifying the number of people who see a campaign with manipulated content or reporting on what foreign countries tried. Very little work investigated how these audiences are affected and whether their views or behavior change as a result. To do so, the team will analyze social media and internet-generated data and will also carry out public opinion surveys and conduct in-depth interviews.

"My incoming hypothesis is that we will find that most online efforts through social media don't have much of an effect,” Shapiro said. “However, we shouldn't discount them entirely because there have been and will continue to be some notable successes."