Study: AI Tool Reduces Political Polarization on X
The AI browser extension lowered hostile political content on X feeds, improving users’ attitudes toward opposing parties by two points, equal to three years of U.S. affective polarization change.
- In a 10-day field experiment, the Stanford-led research team tested an AI-powered browser extension that reorders X feeds to downrank antidemocratic posts with over 1,200 participants during the 2024 U.S. presidential campaign, publishing results Thursday in Science.
- Faced with limited access to proprietary ranking systems, researchers developed a tool to study how social media algorithms steer exposure to antidemocratic attitudes and partisan animosity and to give researchers and end users more control over feed ranking without platform operators' cooperation.
- Survey responses showed participants warmed toward the opposing political party by two points on a 1–100 scale and reported less anger and sadness during use, though emotional effects faded after.
- Researchers argue that platforms could reduce polarization by downranking hostile content, and the team released the tool's code to enable user-controlled algorithms and policy interventions.
- The researchers note the study was limited to browser users and did not measure long-term impact, while authors warn results may not generalize to other platforms and countries.
39 Articles
39 Articles
A team of American researchers has shown that the order in which political messages are displayed on platforms does affect polarization, one of the most debated issues since the emergence of social networks and the breakup that is experienced in much of our societies. The phenomenon is just as strong regardless of the political orientation of the user, academics stand out in an article published today in Science magazine.Social networks are an i…
According to a study, a tool that shifts extremism and violence down can reduce political hostility
New AI tool can lower political temperature and partisan rhetoric through algorithm control
A new Stanford-led research published in Science shows that a web-based tool can drastically reduce partisan animosity on X feeds without platform cooperation. By reordering content to downrank anti-democratic posts, the bipartisan study demonstrated a path toward lowering political polarisation and giving users control over their algorithms New research from Stanford University demonstrates that algorithm intervention can reduce partisan animos…
Study shows that small changes in the tone of posts in X's "for you" feed increase negative feelings toward political opponents
Coverage Details
Bias Distribution
- 43% of the sources lean Left, 43% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium

















