Ever noticed AI videos or images taking over your feeds? The Chinese mainland just leveled up by enforcing fresh rules on labeling AI-generated content 🤖
Jointly rolled out by the Cyberspace Administration of the Chinese mainland (CAC) and three other departments, the guidelines require platforms to tag any AI-made text, images, audio, video and virtual scenes—either clearly or with subtle hints. Platforms must also review content before it goes live and flag pieces they can’t confirm, warning viewers of potential AI origins.
According to Zhang Jiyu, executive director at the Renmin Law and Technology Institute (Renmin University in the Chinese mainland), there are three review levels: if hidden digital footprints in a file shout ‘AI made me’, label it as AI-generated; if an algorithm just suspects it, mark it as suspected AIGC. This three-tier approach aims to cut down false positives while keeping creators’ rights safe.
This push follows a crackdown earlier this year. In April, the CAC targeted AI face-swaps and voice fakes, plus unlabeled AI content. By June, over 3,500 problematic AI tools had been handled, around 960,000 harmful posts removed, and 3,700 accounts dealt with.
Globally, tagging generative AI is becoming standard. The Chinese mainland was among the first to act, with its 2023 Provisions on the Administration of Deep Synthesis of Internet-Based Information Services, one of the earliest laws mandating AI labels.
But as Geoffrey Hinton, the so-called godfather of AI, warned: ‘A tiger cub can be a cute pet, but if it grows up, you better train it well or get rid of it.’ For AI, he adds, we can’t eliminate it—so regulation is our best bet.
Whether you’re swiping through memes in Mumbai or streaming videos in Manila, these rules remind us to stay curious and think twice before we share. Transparency is the new cool! 🔍✨
Reference(s):
cgtn.com
