This report tests how X moderates anti-Muslim dehumanization and violent threats, revealing inconsistent enforcement of its own policies.

Anti-Muslim bigotry and incitement surged across social media in 2026, alongside a sharp rise in hate crimes and attacks on mosques, including the May 18 shooting at the Islamic Center of San Diego that killed three people. This report examines how X (formerly Twitter) handles the most explicit forms of that hostility: posts that dehumanize Muslims as vermin, parasites, or an infestation, and posts that directly call for Muslims to be killed and mosques to be burned.

The findings reveal significant gaps in both proactive and reactive moderation. Content that openly violates X’s own policies, with threats of violence, remained live and accessible for months. Even after it was reported, most of it stayed online.

Posts from paid X Premium accounts were removed less often than posts from standard accounts, and Premium subscribers were more than twice as likely to post violative content repeatedly. With the 2026 midterm elections approaching, the report calls on X to enforce its existing Hateful Conduct and Violent Content policies consistently and to disclose how its proactive detection works. It also urges X to hold verified repeat violators accountable, make user reporting more transparent, adopt surge protocols around elections and other trigger events, and support counterspeech alongside enforcement.

The report is published by the Center for the Study of Organized Hate (CSOH) in partnership with The Bridge Initiative at Georgetown University.

Read the full report here.