《TAIPEI TIMES》MODA working on better social media AI disclosure

The proposed system for identifying AI-generated content begins with creators disclosing its use by checking a box during upload. If this initial step is missed, platforms are expected to automatically label content when signs of AI technology are detected. As a third layer, the ministry plans to establish channels for both government and public reporting of unlabeled AI-generated material that might bypass the first two checks.

This multi-layered approach addresses the challenge of ensuring accurate labeling despite potential creator oversight or attempts to remove embedded AI markers. The ministry acknowledges that some content may slip through automated systems, making public and governmental reporting crucial for adding labels and taking action against content that violates platform rules, especially if used in scams or rights infringements.

Beyond the labeling system, MODA is also advocating for more prominent display of AI content labels, citing YouTube's current placement as insufficiently visible. While Google is reportedly making swift progress, Meta is still engaged in technical adjustments. The ministry also plans to promote AI literacy education to equip individuals with the skills to recognize AI-generated content independently, particularly for publicly available content on platforms like Line, which primarily handles private communication.