MANILA — As AI-generated videos and photos overwhelm social media, video-sharing giant TikTok says it is taking steps to let users take back control and limit the amount of AI "slop" that lands on their feed.
TikTok said the move is also meant to give users greater control over their content recommendations. Doing this is simple, the company said.
On your profile, click Settings and Privacy, then click Content Preferences.
From Content Preferences, click Manage Topics. From there, users can choose how much AI-generated content and even other topics they are willing to see.
The rise of cheap AI video generation tools has recently led to many creators relying on these tools to churn content, instead of shooting authentic videos. It has also spawned concerns about deepfakes and misinformation.
As AI-generated content becomes easier to produce at scale, detection and platform integrity measures are increasingly important to help protect users from misleading or low-quality content, the company said.
TikTok says it has labeled more than 3 billion videos globally as AI-generated through creator labeling tools and invisible watermarking technology.
TikTok also said it took down more than 86 million fake accounts globally during the first 3 months of 2026 for using AI to mass-produce spam.
The company said it is improving its detection systems to identify accounts that post AI-generated spam.
In 2025, TikTok said that of the 1.6 million videos flagged daily for violating TikTok's community guidelines, 85 percent are removed automatically by its AI systems before even reaching the platform's 40,000 trust and safety teams.
TikTok said its AI machines have learned to recognize harmful objects and understand facial and body structures.
Amusing or addictive? The algorithms powering our social feeds
Leaders of major AI firms recently told the UN Security Council that the technology calls for heightened vigilance. OpenAI CEO Sam Altman urged governments to work with each other and the industry on decisions shaping the technology's future.