Overview
- The test adds an option in Settings under Manage Topics called “AI generated content,” letting users choose to see more or less AI videos in their feeds.
- TikTok says about 1.3 billion videos on the platform carry AI labels, though AI content remains a small share of the more than 100 million daily uploads.
- Platform rules require creators to mark qualifying clips as “realistic,” prohibit harmful deepfakes, and apply an “AI-made” watermark to content produced with TikTok’s own tools.
- TikTok plans £1.5 million in funding for partners such as Girls Who Code to produce resources on responsible AI use.
- The company is cutting 439 trust-and-safety roles in London, as it touts automation that reduced graphic content seen by human moderators by 76% over the past year and draws concern over reduced human oversight.