Online information manipulation operations - the transmission of deceiving messages that serve a purpose of the sender - continue to be a growing concern to democracies around the world. It is often assumed that all countries are targeted by information manipulation operations. But is it the case? How does France fit into wider international manipulation operations?
The public tends to learn of information operations during their peaks in activity such as crises, protests, or elections. Research has shown that some information operations against various democracies have been growing and forming for up to a decade before they were identified and suspended on social media platforms. In some cases, this long game of information manipulation attempts by state-backed inauthentic accounts has been found to relate to higher degrees of polarization in conversations online, where people’s attitudes and opinions diverge toward extremes.
This research, conducted by our Visiting Fellow Alexandra Pavliuc, follows the publication of our report Media Polarization "à la française"? Comparing the French and American Ecosystems (2019) and of our policy paper Information Manipulations Around Covid-19: France Under Attack (2020). It aims at establishing a lay of the land, by identifying how France has fit into the wider strategies of state-backed information manipulation networks that have been identified and taken down by Twitter since October 2018. We conducted a broad analysis of all 211 million tweets that Twitter identified and released as part of their information operation (IO) reporting from 19 countries or groups. France-related content was a small portion of most operations, with the highest percentage of France-related content being 8.6% (amounting to 96,839 tweets) in one Iranian operation, and no more than 1.2% of any entire operation (amounting to 120,706 tweets).
We also conducted further qualitative case studies on five operations that contained over 20,000 France-related tweets: Iran, the Internet Research Agency also referred to IRA (Russia), Venezuela, Serbia, and Turkey. Noteworthy countries, such as China, were not found to be targeting France in their operations on Twitter, though this finding should not be generalized to other influence methods. Our analysis showed that these countries engaged in community building content before going political, and published tweets that were framed as news headlines to increase the trustworthiness of their messages. We argue that future research should account for these types of apolitical audience building.