TikTok has announced a series of initiatives aimed at helping users better identify and understand content generated by artificial intelligence, as the social media platform responds to the rapid growth of generative AI.
As part of this effort, TikTok is partnering with the News & Media Literacy Alliance (NAMLE) and deepfake researcher Henry Ajder to develop educational resources aimed at helping users understand how AI technology works and recognize AI-generated content.
TikTok’s AI literacy program, launched in November 2025, has already garnered more than 200 million views through educational content produced by partners such as No Filtr and Raspberry Pi, according to a press release. The company said it has invested more than $4 million in this initiative and plans to expand to reach additional communities around the world.
The platform also announced new measures against AI-generated spam, saying that while advances in generated AI have created new opportunities for creative expression, they have also made it easier for misleading or low-quality content to be produced in large quantities.
TikTok announced that it removed more than 86 million fake accounts in the first quarter of this year under its existing anti-spam policies. Going forward, we will begin testing an enhanced detection system designed to identify accounts primarily used to distribute AI-generated spam, particularly in areas where misleading information can have broad societal impact, such as politics and current events, financial advice, and health information.
TikTok also announced that it has joined the steering committee of the Coalition on Content Provenance and Authenticity (C2PA), an industry group that develops technical standards to help identify AI-generated and digitally altered content. The platform labels more than 3 billion videos using a combination of content credentials, creator-applied labels, and proprietary invisible watermarking technology, using AI to help users recognize when content was created or significantly modified.
