AI Policy Radar

    CN Content Labeling

    China moved first and hardest on synthetic-content provenance; the EU (Aug 2026) and California followed. Global platforms and model providers now face three overlapping labeling regimes — provenance infrastructure has become mandatory plumbing.

    What it is

    From September 2025, all AI-generated content in China must carry explicit labels (visible) and implicit labels (metadata), with platforms required to verify and re-label distributed content — the world's most comprehensive content-provenance mandate in operation.

    Who is affected

    China moved first and hardest on synthetic-content provenance; the EU (Aug 2026) and California followed. Global platforms and model providers now face three overlapping labeling regimes — provenance infrastructure has become mandatory plumbing.

    Key obligations

    • Explicit labels on AI-generated text, images, audio, video and virtual scenes
    • Implicit metadata labels embedded in files
    • Distribution platforms must check and mark content status

    Recent signals

    • 17 July 2026

      China releases international AI ethics governance action plan at WAIC 2026

      Building on its 2025 Global AI Governance Action Plan and the proposed World AI Cooperation Organization, China extends its bid to shape multilateral AI ethics norms.

    • 1 September 2025

      China's AI-content labeling measures take effect

      All AI-generated content must carry explicit and implicit labels, with platform verification duties — the world's most comprehensive provenance mandate in operation.

    • 26 July 2025

      China proposes World AI Cooperation Organization at WAIC

      The Global AI Governance Action Plan and a Shanghai-based multilateral body mark China's most ambitious governance-export move.

    Related instruments

    ← Back to the AI Policy Radar