Political Misinformation.
Publications:
“Generative propaganda: Evidence of AI’s impact from a state-backed disinformation campaign” (PNAS Nexus, 2025) with Carl Ehrett, Darren Linvill, & Patrick Warren.
This paper uses the identification of an instance of generative AI adoption by a Russian-affiliated propaganda site to show how AI-assisted tools significantly increase the volume, breadth, and persuasiveness of disinformation, thereby amplifying state-backed propaganda campaigns.
“ElectionRumors2022: A Dataset of Election Rumors on Twitter During the 2022 U.S. Midterms” (Journal of Quantitative Description, 2025) with Joseph Schafer, Kayla Duskin, Stephen Prochaska, S., Anna Beers, Lia Bozarth, Taylro Agajanian, Mike Caulfield, Emma Spiro & Kate Starbird.
This paper presents an analysis of a dataset comprising 1.81 million Twitter posts on 135 rumors during the 2022 U.S. midterm elections, incorporating exploratory analyses, comparisons with 2020 election rumors, and a mixed-methods case study of Arizona to suggest future research directions on online misinformation and disinformation.
“Social Truth Queries as a Novel Method for Combating Electoral Misinformation: Evidence from Kenya” (International Journal of Press/Politics, 2024) - with Maddy Jalbert
Using data from over 4,800 Kenyan respondents and three survey waves, this project will include several papers related to combatting the spread of misinformation narratives related to fraud and violence stemming from the 2022 Kenyan General Election.
Media Coverage: GW Today.
“Combining Interventions to Reduce the Spread of Viral Misinformation" (Nature Human Behavior, 2022) - with Joseph Bak-Coleman, Ian Kennedy, Andrew Beers, Joseph Schafer, Emma Spiro, Kate Starbird and Jevin West
This paper tests the impact of several disparate methods for combatting online misinformation using a generative model of the spread of viral misinformation. Applied to a corpus of 10.5 million tweets, analyses show that popular intervention techniques are often ineffective in isolation. However, when combined these methods are shown to substantively reduce the spread of misinformation.
“Repeat Spreaders and Election Delegitimization: A Comprehensive Dataset of Misinformation Tweets from the 2020 U.S. Election" (Journal of Quantitative Description, 2022) - with Ian Kennedy, Andrew Beers, Kolina Koltai, Joey Schafer, Paul Lockaby, Michael Caulfield, Michael Grass, Emma Spiro, Kate Starbird and Isabella Garcia-Camargo
This paper details the development and release of a new dataset containing over 49 million tweets and 450 distinct stories linked with misinformation during the 2020 U.S. Election. The paper also uses the contained data to detail the influence of “repeat spreaders” during the election along with asymmetries in the partisanship of users sharing misinformation during the election period.
Media Coverage: New York Times. January 6th Testimony. [Public Data]
“Auditing Google’s Search Headlines as a Potential Gateway to Misleading Content: Evidence from the 2020 U.S. Election" (Journal of Online Trust and Safety, 2022) - with Himanshu Zade, Yuanrui Zhang, Kate Starbird, Ryan Calo, Jason Young and Jevin West
Using more than 800k headlines from Google’s search engine results pages (SERP), we present results from a qualitative coding effort of 5,600 headlines. We show how delegitimizing content was most common among video results while also detailing which search terms and domains were most influential in the dissemination of delegitimizing headlines. We conclude with policy recommendations related to data transparency.
Media Coverage: Tech Policy Press. CIP Summaries. [Public Data]
“Social truth queries: A new user-driven intervention for countering online misinformation” (Journal of Applied Research in Memory and Cognition, 2023) - with Madeline Jalbert, Pragya Arya, and Luke Williams.
Drawing from the work of researchers out of UCT’s Centre for Analytics and Behaviour Change (CABC), this project utilizes both laboratory and real-world evidence to analyze the impact of novel techniques focused on centering perceptions of truth as an alternative method for fighting the spread of misinformation.
Media Coverage: CIP Blog.
“Legislating Misinformation: Evidence of Amplification from Ballot Processing Laws during the 2020 U.S. Election" (Under Review) - with Joey Schafer, Ian Kennedy, Andrew Beers, Rachel Funk Fordham, Emma Spiro and Kate Starbird
This paper uses differential legislation across states to assess how restrictive ballot processing and submission guidelines amplified the spread of localized misinformation during the 2020 U.S. Election. The paper draws from the ElectionMisinfo2020 Database, with prominent misinformation narratives tied to specific states throughout the primary and general election periods.
Papers and Reports:
“The Long Fuse: Misinformation and the 2020 Election” (EIP, 2021) - with partners from Stanford, the DRF Lab, and the Atlantic Council
Alongside researchers from several organizations, I worked with the Center for an Informed Public to identify, report, track, and analyze misinformation arising during the 2020 US Election. The insights of this project have been condensed into the linked report.
“Uncertainty and Misinformation: What to Expect on Election Night and Days After” (EIP, 2020) - with Kate Starbird, Michael Caulfield, Renee DiResta, Jevin West, Emma Spiro, Nicole Buckley and Rachel Moran
This piece drew on available evidence and previous elections to prepare policymakers and researchers for expected misinformation stories in the build-up to the 2020 US Election.
“Emerging Narratives Around ‘Mail Dumping’ and Election Integrity” (EIP, 2020) - with Ian Kennedy, Andrew Beers , Kolina Koltai, Joey Schafer, Paul Lockaby, Michael Caulfield, Michael Grass, Emma Spiro, Kate Starbird and Isabella Garcia-Camargo
This piece details the specifics of a misinformation narrative involving “mail dumping” and its framing as a threat to electoral integrity online.
“Inconsistencies in state-controlled Media Labeling” (EIP, 2020) - with Nicole Buckley, Joey Schafer and Martin Zhang
This analysis details the extent to which major social media platforms labeled government-affiliated media outlets in the build-up to the 2020 U.S. Election.
EIP Media Coverage: The New York Times. The Guardian.
Future Papers:
“Election Observation, Misinformation, & Election Integrity: Evidence from ELOG Communications during Kenya’s 2022 General Election” (Forthcoming) - with Kevin Mudavadi and Maddy Jalbert
Drawing on survey data from Kenya’s 2022 General Election, this book chapter details how the communications of election observation groups can serve to counter both potentially harmful misinformation and justifications for electoral violence.