YouTube quietly launched privacy rules last month aimed at helping users take down AI content that simulates their face or voice.

YouTube creators and users can now request to erase artificial intelligence-generated content meant to look or sound like them. The new policy is not listed under YouTube’s Community Guidelines but instead under its privacy policy, suggesting that simulating someone with AI technology is a violation of user privacy.

“As this type of content becomes more common, it’s important that we offer our community the tools to report when their image or voice appears on YouTube,” the video-sharing platform wrote in a blog post.

However, YouTube will consider a wide range of factors before it issues a takedown notice to a creator over AI-generated content. Content that simulates someone else could be protected under the policy if it’s clear that the video is a parody or satire, if it features a well-known public figure or if the video doesn’t feature the subject endorsing violence or a political cause.

If YouTube finds that the content violates the AI privacy policy, the video creator must remove the video from the website. This is unique for YouTube policy, which usually lets creators change their videos to private, even if they violate YouTube’s Community Guidelines.

The new AI policy comes as tech companies struggle to create new rules around the emerging technology. YouTube and other platforms already require users to label AI-generated content, and some have experimented with automatic label generators to detect when a video or post is made with AI.

• Vaughn Cockayne can be reached at vcockayne@washingtontimes.com.