MIPs: Modeling Influence Pathways

Influence pathways are the dynamic means by which coordinated influence messaging flows across online and traditional broadcast (“offline”) media platforms and communities. For example, in the 1980s, disinformation that AIDS was a U.S. bioweapon spread from less reputable newspapers to more reputable ones as interest in the story grew. 

Similar narratives emerged and propagated at a much larger scale and speed during the Ebola and COVID pandemics of 2014 and 2019, respectively, over newspapers, broadcast, and social media. False narratives reemerged in early 2022, with a variety of actors (e.g. foreign states and fringe groups) propagating the false information across a number of online and offline platforms and communities.

Understanding the pathways by which different types of information propagate across the information ecosystem is important. In addition, identifying patterns that emerge among these pathways, such as the propagation from certain niche platforms to more mainstream platforms, is essential to understanding influence operations.

Modeling Influence Pathways (MIPs) explores artificial intelligence (AI) technologies for the following:

• Connecting various identified influence messaging flows across platforms
• Learning, mapping, and modeling which pathways are used by what types of information
• Discovering patterns that characterize these pathways

MIPs complements maturing capabilities for identification of misinformation, disinformation, and manipulated information with better mapping and understanding of the pathways used to disseminate and amplify that information. These pathways are not static but rather adapt to changes in the information ecosystem in response to emerging platforms, increasing platform moderation or censorship, or new types of messaging. 

Today, pathway discovery is often a manually intensive, retrospective process that fails to match the speed and scale of the information ecosystem. More automated, accurate, and timely identification of influence pathways will enable earlier anticipation and detection of new influence messaging flows; and better forecasting of the impact of changes to the ecosystem (e.g., platform policies) on information flows.