TikTok announced on Thursday its decision to integrate a technology aimed at discerning images and videos created by artificial intelligence (AI) on its platform.
The initiative seeks to introduce greater transparency to content creation processes by adopting “Content Credentials,” a digital watermark developed by Adobe, which signifies the manner in which images were generated and edited. Notably, this technology, while pioneered by Adobe, is available for utilization by other companies and has already been embraced by entities such as OpenAI, the developer of ChatGPT, Reuters news report said.
The move comes amidst growing concerns among researchers regarding the potential misuse of AI-generated content, particularly in the context of influencing the upcoming U.S. elections. TikTok, which earlier this year joined a coalition of 20 tech firms committed to combatting such misuse, is among the platforms taking proactive steps to address this issue. Similarly, YouTube, under Google’s parent company Alphabet, and Meta Platforms, which oversees Instagram and Facebook, have also expressed intentions to leverage Content Credentials.
For the system to be effective, both the creator of the AI tool responsible for generating content and the platform where the content is distributed must consent to utilizing this industry standard. Consequently, if an image is generated using OpenAI’s Dall-E tool, for instance, a watermark is automatically affixed to the resulting image. Upon uploading such a marked image to TikTok, it will be duly identified as AI-generated.
TikTok, a subsidiary of China’s ByteDance, boasts a user base of 170 million in the United States. However, the platform faces regulatory challenges, including recent legislative efforts mandating ByteDance’s divestment of TikTok or potentially facing a ban. TikTok and ByteDance have contested this legislation, asserting that it infringes upon First Amendment rights.
While TikTok already implements labeling for AI-generated content created within its app, this latest measure extends to content originating externally. Adam Presser, TikTok’s head of operations and trust and safety, underscored the platform’s commitment to removing unlabelled realistic AI-generated content, emphasizing adherence to community guidelines and policies.
