IGP Releases New Dataset of Online Foreign Information Operations Targeting Elections
MARGOT FULDE-HARDY*
CAMILLE FRANÇOIS*
In 2016, large-scale information operations on social media orchestrated by the Russian Internet Research Agency, a troll farm based in St. Petersburg, shook the US presidential election, catalyzing new fields of study and practices on the modern practices of foreign interference. Since then, a number of sophisticated campaigns targeting elections have been exposed around the world, yet important questions fundamental to the future of democracy remain:
- How prevalent and consistent are the tactics used throughout these campaigns?
- Have their methods changed or advanced over time?
- Since the first incident using generative AI to target elections in 2019, how has this trend evolved, particularly amid OpenAI's recent disclosure that a number of information operations used their services? And finally, can we identify trends by examining the frequency and volume of these disclosures over time, looking at the nature of the threat and on technology companies and public institutions' disclosures of these incidents?
These questions remain difficult to answer because researchers lack a unified view and analytical framework to make sense of these incidents. To help put these campaigns in perspective and to encourage further research on information operations online, the Institute of Global Politics is releasing a new and comprehensive dataset of online foreign information manipulation and interference (FIMI) spanning 2014 to 2024 and specifically targeting elections. Expanding on existing databases, this database will contribute to developing cross-campaign analysis through the systematic application of a common analytical framework on the foreign information operations exposed. The dataset will help researchers test this methodological approach through using the DISARM Framework, which is one of the available frameworks applying a cybersecurity approach to enumerate tactics, techniques and procedures (TTPs) in the context of online information operations.
We hope that this work will help researchers, journalists, and policymakers contextualize this issue throughout the EU elections of 2024, as well as encourage new research into the evolution of these threats and kick off debate on shared methodologies and best practices that cross disciplines to anticipate and hopefully prevent these information operations from disrupting the democratic process. Please reach out to [email protected] with any feedback or questions.
We thank Tamar Mitts, Assistant Professor of International and Public Affairs, Columbia SIPA; Shelby Grossman, Research Scholar, Stanford Internet Observatory; Alicia Wanless, Director of the Partnership for Countering Influence Operations, Carnegie Endowment for International Peace; Michael Berk, independant consultant; Kevin Limonier, Associate Professor in Geography and Slavic Studies, Université Paris 8; and Ben Nimmo, Principal Investigator, OpenAI, for their extraordinarily valuable feedback on this work.
Available to download:
*Margot Fulde-Hardy is a former researcher at the French Agency in charge of countering foreign information manipulation (VIGINUM), and Investigator at Graphika. This research was conducted with the Institute of Global Politics, independent of and prior to her current and former affiliations.
*Camille François is a Lecturer of International and Public Affairs and IGP Affiliated Faculty at Columbia SIPA