AI may dominate TikTok feeds these days, but the platform is finally handing users a dial to turn the volume up or down.
TikTok says it is testing a feature that lets people control how much AI-generated material they see, after disclosing that its app now contains more than 1.3 billion labelled AI videos.
The trial comes as tools such as OpenAI’s Sora and Google’s Veo 3 unleash a surge of synthetic clips across the internet. Some of the fastest-growing YouTube channels produce nothing but machine-made videos, much of it the low-effort “AI slop” that critics say clutters feeds with surreal or nonsensical content.
Jade Nester, TikTok’s European policy lead for safety and privacy, said the aim was to give users “the power to see more or less” AI material depending on preference.
Within the app’s “manage topic” settings, people will be able to adjust the level of AI content just as they can for categories such as fashion, current affairs or dance.
TikTok will also watermark clips created using its own AI tools or detected through the industry-wide C2PA system, and it will continue removing unlabeled realistic deepfakes.
Alongside the changes, the platform is launching a $2 million AI literacy fund to support groups producing educational material on responsible use.