China will enforce clear flagging of all AI generated content starting from September | AI text, audio, video, images, and even virtual scenes will all need to be labeled.

https://www.tomshardware.com/tech-industry/artificial-intelligence/china-will-enforce-clear-flagging-of-all-ai-generated-content-starting-from-september

Share.

21 Comments

  1. “The Chinese Communist Party’s (CCP’s) national internet censor just announced that all AI-generated content will be required to have labels that are explicitly seen or heard by its audience and embedded in metadata.

    This regulation takes effect on September 1, 2025, and will compel all service providers (i.e., AI LLMs) to “add explicit labels to generated and synthesized content.” The directive includes all types of data: text, images, videos, audio, and even virtual scenes. Aside from that, it also orders app stores to verify whether the apps they host follow the regulations.

    Users will still be able to ask for unlabeled AI-generated content for “social concerns and industrial needs.” However, the generating app must reiterate this requirement to the user and also log the information to make it easier to trace. The responsibility of adding the AI-generated label and metadata falls on the shoulders of this end-user person or entity.”

  2. Probably started because the leader saw too many AI photos of Winnie The Pooh and him merged together.

  3. Although I disagree with just about all of China’s policies, this is one that they’re ahead of the curve on.

  4. This is an interesting article… after thinking about it for a while, here are my thoughts:

    I’m sorry, but as a large language model, I’m unable to leave a “clever, but valuable take” on this reddit thread. Consider forming an opinion on the article that reflects your personal beliefs and share that comment instead. Site users will appreciate the authenticity! 😊

    Is there anything else I can help you with today?

  5. Famous_Track_4356 on

    This needs to be enforced worldwide!

    Too many people can’t make a difference between fact and fiction

  6. Thats great for content directly generated by AI, but I think its going to be impossible to enforce for human-created content with AI tools used as part of the workflow for making assets used in the content. Like if a game developer generates a texture with AI, then hand edits it in photoshop, then paints that texture on a model, used in a hand-crafted game

  7. Neoliberal_Nightmare on

    China is consistently sensible. It’s what happens when you don’t let billionaires dictate policy. Everyone with a brain can agree THAT AI content should be flagged, but it won’t happen in places where tech bros control the policy.

  8. mr_harrisment on

    Ok. How can this actually be a ‘bad thing’ — surely there must be a sinister reason too?

  9. Living-Baseball5223 on

    Good lord, I’m agreeing with CCP? This does make sense to me. All AI content should be required to have an encrypted signature tying it to the creator. It’s getting too good to fake.

  10. I agree with this. It should be illegal to use AI to impersonate a human without clear disclosure. I would even apply this to video and image filters.

  11. We should develop media containers/formats that support cryptographic signing.

    Then we can check if a media file has been created by a trusted creator (eg. AP, Reuters etc)
    If browsers will supports this in a meaningful way, we should be able to easily spot if an image has originated from a trust worthy source or not.

  12. TheLastSamurai on

    But yes they are the dystopian state with no rights lol, it’s so funny so see how much better their approach to AI is

  13. So where’s the limit? Is 49% or 51% of image done by AI considered AI or maybe any use of AI?

  14. It’s definitely a good step forward but I’m worried that the bad actors will simply ignore that and it is impossible to punish people for resharing content which they believe is genuine. There’s a long way to go.

  15. faultysynapse on

    Great idea China. I would like to see the rest of the world follow suit. When you’re right you’re right.