China has released a set of guidelines on labeling internet content that is generated or composed by artificial intelligence (AI) technology, which are set to take effect on Sept. 1.

  • some_dude@lemm.ee
    link
    fedilink
    English
    arrow-up
    56
    arrow-down
    4
    ·
    16 days ago

    This is a smart and ethical way to include AI into everyday use, though I hope the watermarks are not easily removed.

    • jonne@infosec.pub
      link
      fedilink
      English
      arrow-up
      6
      ·
      16 days ago

      It will be relatively easy to strip that stuff off. It might help a little bit with internet searches or whatever, but anyone spreading deepfakes will probably not be stopped by that. Still better than nothing, I guess.

      • some_dude@lemm.ee
        link
        fedilink
        English
        arrow-up
        1
        ·
        13 days ago

        You can use things like steganography to embed data into the AI output.

        Imagine a text has certain letters in certain places which can give you a probability rating that it’s AI generated, or errant pixels of certain colors.

        Printers already do something like this, printing imperceptible dots on pages.

    • Cocodapuf@lemmy.world
      link
      fedilink
      English
      arrow-up
      4
      ·
      15 days ago

      I’m going to develop a new AI designed to remove watermarks from AI generated content. I’m still looking for investors if you’re interested! You could get in on the ground floor!

      • BreadstickNinja@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        14 days ago

        I’ve got a system that removes the watermark and adds two or three bonus fingers, free of charge! Silicon Valley VC is gonna be all over this.