What Are the Politics of a Platform? What the Data Says About Content Moderation on X.

When Elon Musk acquired Twitter, he promised expansive freedom of speech, so it was no surprise when he loosened content moderation policy and reinstated thousands of previously banned accounts. No one disagrees with that. A much more difficult question is what effect this had on the platform, and how we could even know with any degree of precision.

Musk and X repeatedly claimed that hate speech went down post-acquisition, and that the platform’s new “freedom of speech, not freedom of reach” approach is working, suppressing hateful content algorithmically even when it isn’t removed. Researchers, civil society organizations, and EU officials pushed back, pointing to study after study showing hate speech went up.

Both sets of claims may be correct. The disagreement may largely reflect a measurement difference: external researchers can only really measure how many hateful posts were produced; Musk claimed to reduce hateful impressions. Whether the platform’s demotion system actually reduced how often users saw hate speech is a question that requires internal data—data that X has never shared and has made it harder to obtain. This article lays out what the available evidence actually shows, and where it stops.