TikTok is bringing in new tools that allow users to decide how much AI-generated content appears in their feeds, while also testing a more secure way to label synthetic videos.
The update places a new feature inside the “Manage Topics” section of the app, giving people the option to increase or reduce the amount of AI-generated clips pushed to their For You feed.
It works in the same space where users already adjust interest areas such as sports or food. TikTok says the feature is a flexible preference tool: “Manage Topics already enables people to adjust how often they see content related to over 10 categories like Dance, Sports, and Food & Drinks.”
This arrives at a time when AI-created video is becoming extremely embedded in online culture. Platforms across the industry now push synthetic media as a central product.
Meta recently released Vibes, its AI-video feed, and OpenAI followed days later with Sora, a dedicated platform for sharing AI-generated clips. Those videos, once posted elsewhere, quickly found their way back onto TikTok.
Many creators are now using the technology to build visuals for topics far beyond entertainment, including long-form history explainers and celebrity narratives.
From TikTok’s point of view, the new feature solves a simple problem, where some people enjoy the flood of AI content, and others don’t. Placing a slider in user settings allows both groups to customise their experience without hiding anything from the platform altogether.
The change will reach users in the coming weeks, but the adjustments do not stop there. TikTok is also moving to strengthen its labelling system for synthetic media. The company already requires creators to label realistic AI content and uses C2PA’s “Content Credentials” to embed metadata into uploads.
As TikTok explains, “We also use a cross-industry technology called C2PA Content Credentials, which embeds metadata into content that lets us, as well as other platforms who use C2PA, know when something is AI-generated.”
The problem is that these tags can disappear when videos are downloaded, edited, or reuploaded somewhere else. To reduce that risk, TikTok is testing a new layer of protection known as “invisible watermarking”, a hidden marker that only the platform can detect.
Over the next few weeks, the tool will appear in videos created with features like AI Editor Pro and in those uploaded with C2PA credentials.
TikTok says the watermark should make it easier to keep labels intact and track changes to content even after it leaves the app.
Alongside these tools, TikTok is putting money into public education around synthetic media. The company has launched a $2 million AI literacy fund involving more than 20 experts across several countries.
Some of the participating organisations include Girls Who Code, which will create content focused on digital safety and understanding AI’s role online.
The company is also expanding its role in industry work, joining new committees within the Partnership on AI and continuing its involvement in the Framework for Responsible Practices for Synthetic Media.
TikTok notes that future updates will continue as standards develop. The company says it has already refined the way labels appear over the past year and will keep adjusting as more tools roll out across the industry.
