Study reveals X's algorithm amplifies outrage content, affecting Democratic users more
Researchers found that X's feed algorithm favors anger-inducing posts and that this pattern appears more often for users who identify as Democrats.
A recent study published in the Proceedings of the National Academy of Sciences investigated how X's algorithm curates content for users. By enrolling 715 Americans who installed a browser extension, researchers collected data from both For You and Following feeds and linked it to self-reported political leanings and value inventories. The findings indicate the system amplifies posts that generate outrage, with reply activity—though under seven percent of total engagement—receiving disproportionate influence in the algorithm's learning loop.
Democrats reported seeing more of this anger-driven material, a disparity the authors attribute either to a larger supply of right-wing posts or to higher engagement with opposing viewpoints among Democratic users. While the study does not pinpoint a single cause, it highlights the broader implications of opaque platform mechanics for public discourse. X has not officially responded, though former head of product Nikita Bier claimed adjustments have been made to curb such content.
Why it matters
The research shows how platform algorithms can shape political exposure and amplify divisive content, influencing public opinion.
In this story
