A fresh analysis of activity on X shows how regular users can influence the removal of misleading posts.
Researchers found that once a Community Note crosses the platform’s 0.4 helpfulness threshold and becomes public, authors delete those posts at a much higher rate. The data shows a clear increase. Posts with public notes are removed about thirty two percent more often than posts that receive only private evaluations.
The team relied on a regression discontinuity design that exploits the Community Notes threshold. Notes just above the cutoff appear publicly while nearly identical notes just below stay hidden. That split allows a clean comparison of similar posts. The researchers applied the approach across a dataset of 264,600 posts that had at least one note. The sample spans two key windows. The first runs through June to August 2024 before the US presidential election. The second covers January to February 2025. The effect is present in both periods.
The data points to reputation pressure as the main driver. When a note becomes public, it signals that the original claim may be inaccurate. Users respond to that social cue, and those with large audiences tend to react faster. Verified accounts delete noted posts at higher rates and often sooner than unverified accounts. Posts that draw heavy engagement also come down more quickly once a public note appears.
Timing plays a role. The survival analysis in the study shows that public notes not only increase the chance of deletion but also shorten the time until it happens among posts that do get retracted. Private notes do not produce the same outcome.
The supplemental material reveals some patterns that hint at opportunistic posting. A small set of tweets show unusually high engagement relative to the size of the author’s audience. Some rely on emotional phrasing or attach hashtags that do not match the text. These groups delete posts at higher rates once public notes appear. The study frames this as partial support for opportunistic behavior rather than firm proof.
Community Notes occupies a different space than many other interventions used by large platforms. Tools like pre bunking, contextual labels, warning cues, nudges, redirect features, visibility reduction, and account sanctions rely more heavily on centralized design or expert review. Community Notes works through users themselves, which makes it non punitive and scalable. The study describes it as a complementary layer, not a replacement for stricter actions when content is harmful.
The results point to a simple mechanism. Public peer corrections can shift behavior without forcing removals. They prompt authors to reconsider their own posts. The researchers argue that this voluntary response may offer a stable way to limit misleading claims in environments where mandatory removal can provoke resistance.
Image: x
Read next: Google Introduces Gemini 3 with Expanded Search and AI Capabilities
Researchers found that once a Community Note crosses the platform’s 0.4 helpfulness threshold and becomes public, authors delete those posts at a much higher rate. The data shows a clear increase. Posts with public notes are removed about thirty two percent more often than posts that receive only private evaluations.
The team relied on a regression discontinuity design that exploits the Community Notes threshold. Notes just above the cutoff appear publicly while nearly identical notes just below stay hidden. That split allows a clean comparison of similar posts. The researchers applied the approach across a dataset of 264,600 posts that had at least one note. The sample spans two key windows. The first runs through June to August 2024 before the US presidential election. The second covers January to February 2025. The effect is present in both periods.
The data points to reputation pressure as the main driver. When a note becomes public, it signals that the original claim may be inaccurate. Users respond to that social cue, and those with large audiences tend to react faster. Verified accounts delete noted posts at higher rates and often sooner than unverified accounts. Posts that draw heavy engagement also come down more quickly once a public note appears.
Timing plays a role. The survival analysis in the study shows that public notes not only increase the chance of deletion but also shorten the time until it happens among posts that do get retracted. Private notes do not produce the same outcome.
The supplemental material reveals some patterns that hint at opportunistic posting. A small set of tweets show unusually high engagement relative to the size of the author’s audience. Some rely on emotional phrasing or attach hashtags that do not match the text. These groups delete posts at higher rates once public notes appear. The study frames this as partial support for opportunistic behavior rather than firm proof.
Community Notes occupies a different space than many other interventions used by large platforms. Tools like pre bunking, contextual labels, warning cues, nudges, redirect features, visibility reduction, and account sanctions rely more heavily on centralized design or expert review. Community Notes works through users themselves, which makes it non punitive and scalable. The study describes it as a complementary layer, not a replacement for stricter actions when content is harmful.
The results point to a simple mechanism. Public peer corrections can shift behavior without forcing removals. They prompt authors to reconsider their own posts. The researchers argue that this voluntary response may offer a stable way to limit misleading claims in environments where mandatory removal can provoke resistance.
Image: x
Read next: Google Introduces Gemini 3 with Expanded Search and AI Capabilities
