YouTube Unveils New Labels for AI-Generated Videos: Explained

YouTube announced today that it will begin rolling out a new requirement for creators to disclose when videos contain artificial intelligence (AI) generated or synthetic content. As part of the new policy, YouTube will display labels on videos to notify viewers if they contain altered, synthetic or AI-generated material. Google says the labels are designed to provide more transparency and build trust between creators and their audiences. 

AI Labels Explained

Starting this week, YouTube creators will now have to select a checkbox when uploading videos to indicate if a video contains "altered or synthetic media" depicting realistic scenes or people. It includes using AI to:

• Replace one person's face with another's and generate a person's voice

• Alter footage of actual events or locations.

• Generate realistic depictions of fictional major events that did not actually occur.

However, YouTube said disclosures will not be mandatory for content involving filters, effects or animation. The creators also do not need to disclose if AI was used for productivity purposes like generating scripts or captions. According to YouTube, the goal is to increase transparency around altered or synthetic content that could mislead viewers while not stifling creativity.

How It Will Work?

The labels will initially roll out on YouTube's mobile app and website in the expanded description, with plans to expand to TV streaming and other platforms; however, for videos on sensitive topics like health, news, or finance, a more prominent label will display directly on the video player itself.

YouTube noted it may also add AI disclosure labels itself if creators fail to properly indicate synthetic content, especially if that content risks misleading viewers.

Why do AI Labels Matter?

Creators Embracing AI: The rise of advanced AI tools like Gemini, ChatGPT, DALL-E, and Midjourney has made synthetically altering video and audio easier than ever before. YouTube's new policy aims to keep viewers informed when AI has been used to depict fictional yet realistic scenarios.

"Generative AI is transforming the ways creators express themselves," YouTube stated, noting AI can assist with ideation, scripting, captioning and more. Many top creators have already experimented with AI tools for tasks like voiceovers, background edits and concept art. The new labelling policy embraces these capabilities while upholding transparency.

Increased Transparency: In its announcement, YouTube cited that viewers "increasingly want more transparency about whether the content they're seeing is altered or synthetic."

YouTube's AI labels are designed to "build trust between creators and their audience" and avoid misleading audiences about whether events or people are genuine. YouTube is also a founding member of the Coalition for Content Provenance and Authenticity (C2PA), which aims to increase transparency around digital objects.

While imperfect, the honor-system-based AI labels strive to keep viewers adequately informed about what they're watching on YouTube.