SAN FRANCISCO: TikTok on Thursday said it will begin labeling AI-generated content automatically from several platforms, including OpenAI’s Dall-E and the company’s own tools and generators to enhance transparency and trust.
This initiative marks a pioneering step in social media platforms’ efforts to address concerns regarding the authenticity and provenance of digital content, particularly in light of the growing threat posed by deepfakes.
Adam Presser, Head of Trust & Safety at TikTok, emphasized the importance of transparency in the rapidly evolving landscape of AI-generated content. He stated, “AI-generated content is an incredible creative outlet, but transparency for viewers is critical.”
The automatic labeling of AI-generated or edited content represents a groundbreaking development in social media authentication protocols. TikTok, owned by Chinese tech giant ByteDance, revealed its intention to introduce a label that will be applied automatically to content identified as edited or created using AI technologies.
According to TikTok, this labeling initiative will encompass content produced through Adobe’s Firefly tool, TikTok’s proprietary AI image generators, and OpenAI’s Dall-E. Additionally, the company announced its participation in a coalition led by Adobe, aimed at promoting industry-wide labeling standards for AI-generated products, commonly referred to as watermarking.
The Coalition for Content Provenance and Authenticity (C2PA) has developed a technical standard to enhance the identification and traceability of AI-generated content. By partnering with this coalition and other industry peers, TikTok seeks to empower creators to responsibly explore AI-generated content while mitigating the dissemination of harmful or misleading content.
Initially, TikTok’s auto-labeling feature will be rolled out gradually. However, as more platforms adopt the C2PA standard for labeling AI-generated content, TikTok aims to expand its labeling capabilities to encompass a broader range of content sources.
Meanwhile, OpenAI, the Microsoft-backed AI company responsible for the creation of Dall-E and ChatGPT, announced its participation in the C2PA initiative. Additionally, OpenAI unveiled a new tool designed to detect AI-generated digital images.
