The latest transparency reports from Meta (Feb 2025), TikTok (Dec 2024), X (April 2025), and YouTube (Feb 2025) vary considerably in completeness and usefulness.

While all reports include the number of human moderators per language, the data shows how limited the moderation capacity is for smaller or less widely spoken languages. X’s Transparency report does not provide a meaningful link between the volume of illegal hate-speech reports received and the number of items removed. Meta aggregates hate speech within a broader ‘illegal content’ category,” and TikTok has almost no data available specifically on hate speech. YouTube provides figures on submissions by Trusted Flaggers, but this is “the exception rather than the norm.

None of the platforms give any information on how removal decisions are made or how they demote content, and none provide a breakdown of hate-speech enforcement by country or by protected characteristic.

Read the full report here.