Social media giant Meta is set to make significant adjustments to how it labels AI-generated content on its platforms Instagram, Facebook, and Threads. Starting next week, the "AI Information" label for content edited or modified by AI tools will be moved to the post menu, no longer directly displayed under the user's name.
This change brings several key points:
1. Label Distinction: Content fully generated by AI will still retain a prominent label, while labels for content merely edited by AI will be hidden.
2. Transparency of Origin: Meta states that the labeling of AI-generated content will be based on industry-shared signals or user self-disclosure.
3. Reason for Adjustment: The company claims this move is intended to "better reflect the extent of artificial intelligence used in its platform content."
However, this policy adjustment has also sparked controversy. Critics argue that hiding the AI label could increase the risk of user deception, especially in the context of increasingly sophisticated AI editing tools.
It's worth noting that this is not the first time Meta has adjusted its AI content labeling. In July of this year, the company changed "AI-Made" to "AI Information" in response to complaints from photographers about mislabeled authentic photos.
Meta's decision reflects the challenges social media platforms face in handling AI content: how to balance transparency with optimizing user experience. With the rapid development of generative AI technology, we can expect more adjustments to related policies in the future.