TikTok introduces new tag standards for AI content, enhancing transparency

TikTok is actively addressing the proliferation of AI-generated content on its platform by introducing new tags to enhance transparency and combat misinformation. This is part of TikTok's extensive efforts to maintain content authenticity and protect user rights.

Recently, AI-generated content has gained popularity on TikTok, with many creators using AI voice cloning technology to create viral content, such as lip-syncing songs and motivational speeches by celebrities.

To regulate this phenomenon, TikTok requires creators to clearly label their AI-generated content. This can be achieved by adding a new "AI Generated" tag below the username in the corner of the video or through other forms of disclosure, such as stickers or subtitles. TikTok also warns that AI content that is not properly labeled may be at risk of being removed.

This requirement is not new for TikTok, as the platform has already required creators using its internal AI tools to implement similar labeling. However, expanding this requirement to include content from other platforms demonstrates TikTok's firm commitment to safeguarding user awareness. In the digital world, the line between reality and synthesis is increasingly blurred, making it crucial to ensure that users fully understand the nature of the content they consume.

To better achieve this goal, TikTok has announced a partnership with the Content Authenticity and Proof Alliance (C2PA) to label AI-generated content (AIGC) uploaded from other platforms. C2PA is an industry alliance led by Adobe, with members including Google, Meta, OpenAI, Microsoft, and Intel. The alliance provides technical standards for the source and history of media content, aiming to address misleading information on the internet.

TikTok will adopt C2PA's "Content Credentials" technology to attach metadata to images and videos, clearly identifying whether the content is AI-generated. This includes AI content created using external tools such as DALLE and Firefly. According to TikTok's announcement, this labeling feature has already been applied to images and videos, with plans to expand it to audio-only content in the future. This makes TikTok the first video-sharing platform to adopt C2PA's content credentials technology.

In addition, TikTok has joined the Content Authenticity Initiative (CAI) led by Adobe, actively promoting the application of Content Credentials across the industry. Dana Rao, General Counsel and Chief Trust Officer at Adobe, appreciates TikTok's decision and emphasizes the importance of providing the public with methods to discern authenticity in the age where digital content can be easily manipulated.

Lastly, TikTok will increase its investment in media literacy by releasing educational videos and resources to help users understand the importance of these labels and the reasons behind them. To achieve this, TikTok has partnered with MediaWise to develop a series of videos that focus on media literacy skills and the role of AI labels.

Overall, TikTok's implementation of C2PA's content credentials is a significant step towards transparency in the era of generative AI. By providing clear labels and background information, users can make more informed judgments about the content they engage with, thereby creating a safer and more trustworthy online environment.