Angry Online? Stanford Scientists Have A Tool That Could Detox Your Newsfeed

Angry Online? Stanford Scientists Have a Tool That Could Detox Your Newsfeed

Last updated: 2025/11/30 at 3:07 PM

News Room Published 30 November 2025

Researchers at Stanford have demonstrated a new tool that enables users to reduce the amount of aggressive politically partisan content they see on social media, without involvement from the platform they’re scrolling. The researchers found that simply reducing exposure to these types of posts was enough to shift out-party partisan animosity by over 2 points on a standard 0–100 feeling thermometer by the end of the study.

According to a new paper first published in Science, the scientists created a web-based tool that automatically reorders the content you see—for example, pushing down posts that contain “partisan animosity,” such as calling for jail time for political opponents or advocating for violent protest. The app did not actually block or hide content.

The study surveyed 1,256 participants who already used X during the weeks leading up to the 2024 presidential election and used the tool for 10 consecutive days. Researchers found no detectable differences in the tool’s impact across party lines, meaning self-identified Democrats and Republicans had their partisan feelings roughly equally impacted by the order of posts in their X newsfeed.

The study also found that the tool could have a discernible impact on the other emotions of the subjects. Anger was the most impacted emotion, showing a slight reduction, followed by sadness, which was also reduced. Positive emotions, like excitement or calmness, were not impacted. Though the study tracked emotions during the 10-day experiment, it didn’t provide any data on whether changes in either partisan feelings or emotional responses remained after the study ended, though the paper suggests this as a topic for future research.

The researchers said that their findings may provide a potential pathway for platforms to address these challenges, such as political polarization and societal division, through “algorithmic interventions.”

Recommended by Our Editors

“Together, these interventions may result in algorithms that not only reduce partisan animosity but also promote greater social trust and healthier democratic discourse across party lines,” read the paper.

Divisive political content may well continue to be largely impossible to avoid online, as many of the most political platforms, like X and Bluesky, have very clear leanings among their respective political influencers.