YouTube introduces automatic labeling of videos created with AI

YouTube has announced that it is beginning to introduce automatic detection and labeling of videos created or significantly altered using artificial intelligence. With this, the platform is moving away from its previous practice, where creators had to declare the use of generative AI tools themselves.

In parallel, the placement of AI labels is also changing – they will be positioned more prominently so that viewers can immediately get context regarding the origin of the content.

Automatic AI content detection

"Starting in May 2026, we are introducing new internal signals to detect content created with the help of AI," YouTube states in its official blog. "If a creator has not disclosed whether AI was used, but our systems detect significant use of photorealistic AI, we will automatically add a corresponding label."

This change fills a gap in the platform's current policy: since 2024, content creators were required to mark themselves when uploading realistic videos based on modified or synthetic media. According to the new rules, authors who believe their video has been labeled incorrectly will be able to correct the disclosure status via YouTube Studio.

AI labels move to the forefront

YouTube is also making visual changes to make AI labels more noticeable. For longer videos, the warning will now appear directly below the video player, above the description. In YouTube Shorts, the label will be overlaid on the video itself.

"By moving these labels to a more prominent location, we allow viewers to get the necessary context immediately," the platform explains. "This is now a unified label format for all photorealistic content, as well as content significantly altered or generated using AI on YouTube."

Content that clearly does not look realistic, is animated, or is only slightly modified will continue to receive a warning only in the expanded description.

Permanent labels for certain videos

For some categories of content, AI labels will be permanent and cannot be removed. This applies to videos created using YouTube's own AI-based tools – such as Veo or Dream Screen – as well as clips containing C2PA metadata, which indicate that the content is entirely AI-generated.

YouTube is a leading member of the Coalition for Content Provenance and Authenticity (C2PA), which developed the C2PA standard for tracking the origin of digital materials. This allows for clearer traceability of the content creation process.

No impact on recommendations and monetization

The platform emphasizes that the mere presence of an AI label does not affect recommendation algorithms and does not disqualify a video from monetization. The labeling is informative in nature and aims to increase transparency, rather than being a punitive measure against creators.

The changes come amid growing pressure on major platforms from regulators around the world – including in the context of the EU AI Act, which from August 2026 introduces transparency requirements regarding content created with the help of AI. YouTube's new steps are a sign that the company is trying to stay ahead of regulations and build standards for disclosing AI use at scale.