Meta plans to label AI-generated content from May 2024

Discontinue removal of content based on its manipulated video policy

Meta also intends to extend labeling beyond manipulated content

Meta plans to commence labeling AI-generated content in May 2024, while discontinuing the removal of content solely based on its manipulated video policy come July. Over the past four years, and notably in the recent year, various forms of realistic AI-generated content such as audio and photos have emerged, with the technology rapidly advancing.

Monika Bickert, Vice President of Content Policy at Meta, emphasized the importance of addressing manipulative content portraying individuals engaging in actions they never actually performed.

Also read: Youth behind Rashmika Mandanna’s recent deepfake video arrested

She mentioned that labels like “Made with AI” on AI-generated video, audio, and images will rely on industry-shared signals of AI images or individuals voluntarily disclosing the use of AI-generated content during upload. Bickert noted the existing practice of adding “Imagined with AI” to photorealistic images produced using Meta’s AI feature.

The decision to label AI-generated content in May and halt the removal of manipulated content based solely on the policy in July aims to familiarize users with the self-disclosure process before ceasing the removal of such content, according to Bickert.

Additionally, Meta intends to extend labeling beyond manipulated content as recommended by the Oversight Board. Bickert mentioned in a blog post that if digitally-created or altered images, video, or audio pose a significant risk of materially deceiving the public on important matters, more prominent labels may be added to provide users with enhanced information and context.