China: TC260 adopted Guidelines on Identification Methods for AI-Generated Synthetic Video Content

On 28 August 2025, the National Cybersecurity Standardisation Technical Committee (TC260) adopted the Cybersecurity Standard Practice Guide — Artificial Intelligence Generated Synthetic Content Identification Method: File Metadata Implicit Identification for Video Files, formulated to implement the Artificial Intelligence Generated Synthetic Content Identification Measures in accordance with the mandatory national standard GB 45438-2025 Cybersecurity Technology — Identification Methods for Artificial Intelligence Generated Synthetic Content. The Guide specifies implicit metadata identification methods for artificial intelligence generated synthetic video files, providing technical schemes across multiple formats, including MP4, MOV, AVI, MKV, WebM and FLV, as well as an extensible metadata platform (XMP) scheme for other formats. It sets out embedding methods through native structures or XMP namespaces, detailing storage locations, key identifiers, and serialised values incorporating mandatory fields such as “Label”, “ContentProducer”, “ProduceID”, “ContentPropagator” and “PropagateID”. The Guide includes normative references to GB 45438-2025 and international standards such as ISO/IEC 14496-12, ISO/IEC 14496-14 and ISO 16684 series on XMP, alongside appendices providing implementation examples for each video format using tools such as ffmpeg and exiftool, as well as code-level development references for integration in RIFF-based containers. The scope applies to artificial intelligence video content generation service providers and internet information video content propagation service providers, with the objective of guiding consistent adoption of implicit metadata identifiers in video content to ensure compliance with national cybersecurity regulatory requirements.