Artificial intelligence chatbots, which learn a lot of information posted on the Internet and provide answers, have always been controversial about copyright infringement. In fact, related lawsuits are in progress in various fields, including music, images, publishing, and coding.
Along with legal action, copyright protection technology is also advancing. Technology is being developed to recognize and block artificial intelligence chatbots when they attempt to collect information, and transparent watermark technology that is not visible in the original image but can only be identified by artificial intelligence has also emerged.
What stands out even more is the digital dock technology. When creators post their work on the Internet, if they plant toxic code, the poison spreads to the artificial intelligence model that collected it and affects other data. The current toxicity results are displayed on the screen. At the top, the desired order name and general results are displayed. The second row is the result of adding 50 poisons, giving the animal an awkward face and six legs. And the third line shows a combination of a dog and a cat when 100 were added, and in the fourth line, an unexpected picture of a cat was created when 300 were added.
These technologies are evaluated for their potential to destroy artificial intelligence models beyond copyright protection. For the time being, copyright issues regarding artificial intelligence learning data will continue, but the battle between spear and shield will continue, and we will see the outcome.