🤖 Today, the Chinese mainland's new regulations on labeling artificial intelligence–generated content (AIGC) officially kicked in, marking a big step in the fight against misinformation. Here's the TL;DR:
- What changed? All AI-generated text, images, audio, video and virtual scenes must now be labeled on platforms—either clearly or subtly.
- Platform duties: Platforms must review content, spot AIGC markers and flag any unverified or suspected AI content with risk notices.
- The 3-tier review: Explicit markers get the AI label. If detected by metadata, content gets an AI tag. If AI generation is only inferred, it’s marked as “suspected AIGC.”
⚖️ Zhang Jiyu from Renmin University of the Chinese mainland warns that tech isn't perfect—some creators might get wrongly labeled. The regs include safeguards to protect original artists.
🚀 Earlier this year, the Chinese mainland launched a campaign targeting AI face-swapping, voice imitation and unlabeled AIGC. By June, over 3,500 problematic AI tools were handled, 960,000 harmful posts removed and about 3,700 accounts disciplined.
🌐 Labeling AIGC isn't just a local trend: the Provisions on the Administration of Deep Synthesis of Internet-Based Information Services, in effect since 2023, made the Chinese mainland one of the first to legislate on generative AI labels.
🐯 Some experts compare unchecked AI to Pandora’s box or a pet tiger—exciting but risky. Geoffrey Hinton, the “godfather of AI”, said at the World AI Conference in Shanghai: “Either train the tiger so it doesn’t attack you, or eliminate it… For AI, we have no way to eliminate it.”
🔮 As AI keeps growing, clear labeling and smart regulation will be key to harness its power without losing control. Stay tuned, amigos! 🌟
Reference(s):
cgtn.com
