A field experiment conducted on the social media platform X, formerly Twitter, revealed how an algorithmic feed influences user behavior and political attitudes compared to a chronological feed. The findings, which were reported in an article published in Nature on 18 February 2026, specifically showed that users tend to shift their political views and options toward more conservative positions when using an algorithmic feed.
Favoring Conservatism New Evidence Reveals that the Algorithm of X Not Only Promotes Right-Wing Activists But Also Turns Users Into Right-Wingers
Background
Previous research on social media platforms Facebook and Instagram, led by J. A. Tucker, found that switching off the algorithm had no detectable political effects on attitudes and behavior during the 2020 elections in the United States. However, in a newer investigation by Germain Gauthier and his team, results showed otherwise, especially on the X platform.
The team conducted an experiment with X users in the United States between 2023 and September 2023. A total of 4965 participants completed the post-treatment survey. Each was randomly assigned to either the algorithmic For You feed or the chronological Following feed for 7 weeks. Users who switched between these two feeds were also examined.
Surveys were provided before and after the 7-week experimental period. These included questions on views regarding policy priorities, the war in Ukraine, and criminal investigations into Donald Trump. Partisanship and affective polarization were also determined. Online behaviors such as network changes and engagement activities were tracked and analyzed.
The team also analyzed more than 260000 posts appearing on the feeds to determine whether the algorithm was disproportionately promoting or demoting specific types of political content or media sources. Unsupervised computational linguistics was used to quantify latent narrative structures and political sentiment in the massive dataset of posts.
Main Findings
• Shift Toward Conservatism
The experiment found that switching users from a chronological feed to an algorithmic feed shifted their political opinions toward more conservative positions. This was most notable regarding policy priorities, perceptions of criminal investigations into Donald Trump, and views on the war in Ukraine.
• Asymmetric Persistence
Moreover, while moving to the algorithm made users more conservative, switching users from the algorithmic feed back to a chronological feed did not reverse the effect. The conservative influence persisted even after the algorithm was turned off.
• Engagement and Content Promotion
It is also worth mentioning that the algorithmic feed increased overall user engagement on the platform but was found to actively promote conservative content while demoting posts and other content from traditional media organizations.
• Following Behavior
The algorithmic feed exposed users to several conservative political activist accounts. This resulted in users following such accounts. Moreover, because users continue to follow these social media accounts even after switching back to a chronological feed, the conservative shift remains, thus explaining why the effects are so persistent.
Takeaways
The shift toward conservative views within the algorithmic feed of X is not a simple byproduct of user choice. The content analysis of hundreds of thousands of posts showed that the algorithm of the platform actively promotes conservative-leaning content while simultaneously demoting posts from traditional media outlets and mainstream news organizations.
Moreover, although the platform prioritizes engagement like any other social media platform, it also favors content from political activists over professional journalism. This could be because these users tend to have more engaging personal brands, or because conservative or right-wing content creators outnumber liberal creators and traditional media.
Gauthier et al. reminded that their findings are specific to X and the time of the research. However, because political discourse is almost always active in X and other social media platforms, and in addition to the growing polarization within the U.S. political landscape, the insights drawn from these findings raise serious questions about platform neutrality.
FURTHER READINGS AND REFERENCES
- Gauthier, G., Hodler, R., Widmer, P., and Zhuravskaya, E. 2026. “The Political Effects of X’s Feed Algorithm.” Nature. DOI: 1038/s41586-026-10098-2
- Guess, A. M., Malhotra, N., Pan, J., Barberá, P., Allcott, H., Brown, T., Crespo-Tenorio, A., Dimmery, D., Freelon, D., Gentzkow, M., González-Bailón, S., Kennedy, E., Kim, Y. M., Lazer, D., Moehler, D., Nyhan, B., Rivera, C. V., Settle, J., Thomas, D. R., … Tucker, J. A. 2023. “How do Social Media Feed Algorithms Affect Attitudes and Behavior in an Election Campaign?” Science. 381(6656): 398-404. DOI: 1126/science.abp936