YouTube has a new rule that lets people ask to remove videos that use fake versions of their faces or voices.
With this rule people can ask YouTube to take down AI-made videos that break their privacy. They don’t need to worry about the video being misleading or a deepfake.
YouTube’s updated Help page explains how to do this. Usually the person asking for the video to be removed has to be the one affected, unless they are a child, do not have access to a computer, are deceased or fall under other special cases.
But asking for a video to be taken down does not mean YouTube will always remove it. YouTube will look at different things like if the video says it is AI-made, if it clearly shows a person or if it is a joke, parody or important to the public.
YouTube will also check if the video shows a famous person and if it shows them doing something sensitive like crime, violence or endorsing a product or political candidate.
When someone complains YouTube will give the video maker 48 hours to remove the video. If they do, the complaint is closed. If not YouTube will review the complaint.
If YouTube decides to remove a video then it will be completely removed from the site.
YouTube’s way of handling AI-made videos is careful because YouTube also uses AI for things like summarizing comments and answering questions about videos.
If someone complains about a privacy issue with an AI video, YouTube will not punish the original video maker. Instead they will focus on fixing the privacy problem which is different from breaking Community Guidelines.
