The YouTube video service announced that from now on, users can manually tag videos with artificial or AI-generated content. Creators will see this checkbox when uploading and publishing their material, and must disclose “altered or synthetic content that appears to be realistic.” These include, but are not limited to, cases where a real person in the video says something they didn’t actually say, or does something they didn’t do, as well as changing footage of real places and events. As an example, YouTube cites a fake tornado heading towards a real city. At the same time, you don’t need to flag “obviously unrealistic content” such as animation, or filters or special effects like blurred backgrounds.

YouTube can add AI labels to videos even if they don’t, because they have the tools, and promise to be extra vigilant about altered or artificial content that could confuse or deceive viewers, mostly on sensitive topics such as health, elections, and finances. However, the software for detecting AI-generated content, according to YouTube spokesman Jack Malone, is very inaccurate, so the platform admits that it relies on the honesty and decency of video creators in this matter.

Літературний редактор.