Memes have become a critical communication tool for partisan and extremist groups. CCPP scholars took a deep dive into the content of memes on alternative (or fringe) platforms, and what they found is just as concerning as it is fascinating. There is no doubt that 2024 will be saturated with online political content, but some of the most dangerous content should be expected to come in “meme” form.
Sponsored by a grant from the Global Network on Extremism and Technology (GNET), a CCPP team employed state-of-the-art deep learning image and visual rhetorical analysis to analyze memes posted on far-right platform Parler. Detailed in the CCPP Report, the team categorized memes into themes of gender, race, partisanship and violence. Researchers then measured the transmission rates of memes associated with each theme, finding, among many other patterns, that violent memes are on the rise. It was also noted that memes with high engagement levels were often branded by a group emblem or logo, increasing their credibility within public discourse.
Under the guise of an unassuming image or inside joke, memes tend to hide their political intent and are more able to escape algorithmic content moderation or filtering. However, the report highlights how moderating memes is critically important for limiting online content that promotes political violence. The methodology and findings of the report provide a working model for meme content filtering that can help platforms identify and filter memes with extremist content.