The 2024 US presidential election created ideal conditions for studying algorithmic polarization. Fake images and AI-generated propaganda proliferated across social platforms, reaching tens of millions of viewers and creating an information environment where truth and fabrication became increasingly difficult to distinguish. Against this backdrop, researchers measured how feed composition affects political attitudes.
Divisive viral content during the campaign included fabricated images designed to damage candidates’ reputations and AI-generated propaganda presenting false narratives. Some posts reached 84 million views, demonstrating the massive scale at which misinformation can spread on modern platforms. This content ecosystem provided researchers with naturally occurring examples of exactly the types of divisive posts they wanted to study.
Over 1,000 users participated in an experiment where researchers manipulated their feeds to show more or less of this divisive content. The timing was intentional—election periods heighten political engagement and emotion, potentially magnifying algorithmic effects. Results confirmed that even during this intense period, subtle feed adjustments produced measurable polarization shifts.
The research revealed that one week of slightly increased exposure to election misinformation and divisive content created polarization equivalent to three years of natural societal change. This suggests that high-stakes political events combined with algorithmic amplification create particularly powerful conditions for rapid attitude shifts.
Understanding these dynamics matters for protecting electoral integrity. If social media algorithms can significantly shift political attitudes during campaign periods, they potentially influence election outcomes beyond whatever effects the content itself might have. This raises questions about whether platform algorithms should be considered part of democratic infrastructure requiring special oversight during elections.
Viral Misinformation During Elections: The Perfect Storm for Polarization
Picture credit: www.universe.roboflow.com

