YouTube lays out new rules for ‘realistic’ AI-generated videos


Many companies and platforms are debating how to manage AI-generated content as it becomes more prevalent. One of the main issues that worries many To clarify whether the AI ​​model produced a photo, video or audio piece. For this purpose, there is for tagging videos made with artificial intelligence.

Starting today, the platform will require anyone to upload “real-looking video made with altered or synthetic media. “Labeling it for the sake of transparency. YouTube defines real content as anything that a viewer could “easily mistake” for an actual person, event or place.

A screenshot of the YouTube Creator Studio, including a question asking the creator if the video contains digitally altered or synthetic content.A screenshot of the YouTube Creator Studio, including a question asking the creator if the video contains digitally altered or synthetic content.

YouTube

If the creator uses a synthetic version of a real person’s voice to narrate the video, or replaces someone’s face with someone else’s, they must add a tag. They should also include a disclosure when they alter footage of a real event or place (for example, by altering an existing cityscape or making it look like a real building is burning).

YouTube says it can apply one of these labels to a video if the creator hasn’t done so, “especially if the modified or synthetic content has the potential to confuse or confuse people.” The team notes that while it wants to give creators some time to get used to the new rules, YouTube will likely penalize persistent policy violators by not including tags when necessary.

These tags will begin appearing on YouTube in the coming weeks, starting with the mobile app, desktops, and TVs. They will mostly appear in an expanded description, noting that the video contains “altered or synthetic content,” adding that “audio or visuals have been significantly edited or digitally generated.”

A screenshot showing how a statement is made A screenshot showing how a statement is made

YouTube

However, when it comes to more sensitive topics (such as news, elections, finance and health), YouTube will place a label directly on the video player to make it more prominent.

Creators won’t need to include a tag if they’re only using generative AI to help with things like scripting, generating ideas for videos, or generating automatic titles. Tags will not be necessary for “obviously unrealistic content” or when the changes are trivial. Using special effects, such as adjusting colors or adding only background blur, will not require creators to use a modified content tag. No lighting filters, beauty filters or other enhancements will be applied.

In addition, YouTube says it is still working on an updated takedown request process for synthetic or modified content. depicts the face or voice of a real, identifiable person. We plan to share more information about this updated procedure soon.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *