YouTube Prepares AI Avatar Tools as Shorts Cross 200B Views

As short-form video platforms compete on speed, scale, and automation, YouTube is preparing to introduce a feature that could redefine how creators show up on screen, without being there at all.

The company plans to allow creators to produce Shorts using their own AI-generated likeness, effectively enabling a digital twin to appear in videos. The announcement came during YouTube CEO Neal Mohan’s annual letter, where he reiterated that “AI will remain a tool for expression, not a replacement.”

While YouTube has not shared a launch timeline or technical specifics, the move signals a deeper push into AI-assisted content creation at a time when platforms are under pressure to balance creator productivity with content quality.

From Camera Time to Compute Time

If implemented as expected, the feature could reduce the friction involved in producing Shorts, particularly for creators who post frequently or operate across formats and languages. The idea mirrors a broader industry shift toward AI-assisted presence, where creators can scale output without being physically present for every recording.

The functionality is expected to resemble systems that create photorealistic avatars from short selfie videos and voice samples, allowing creators to generate videos using their likeness. Such tools replace repetitive recording workflows with a data-driven production layer, an efficiency play that aligns with YouTube’s growing creator economy infrastructure.

For creators managing channels as businesses, especially those focused on education, explainers, or brand partnerships, the appeal is operational: more content, less manual effort.

Disclosure as a Platform Mandate

However, YouTube appears keenly aware of the risks that come with synthetic media at scale.

Mohan emphasised that the platform already labels content generated using its AI tools and requires creators to disclose when they have created “realistic altered or synthetic content”. He also noted that YouTube will “equip creators with new tools to manage the use of their likeness in AI-generated content.”

This focus on disclosure reflects growing concerns around impersonation, misinformation, and viewer trust, issues that intensify when AI-generated humans become indistinguishable from real ones.

Rather than leaving governance to creators alone, YouTube is positioning itself as an active moderator of AI usage, embedding policy controls alongside creative tools.

The ‘AI Slop’ Problem

The announcement lands amid a wider debate around what Mohan described as “AI slop”: low-quality, repetitive, or spam-like content generated at scale.

“Over the past 20 years, we’ve learnt not to impose any preconceived notions on the creator ecosystem,” Mohan said, while adding that YouTube has a responsibility to “maintain the high-quality viewing experience that people want.”

To that end, the company says it is extending its existing systems designed to combat spam and clickbait to also limit the spread of low-quality AI-generated content.

This suggests that AI avatars may increase output, but not all content will be treated equally by the algorithm—placing renewed emphasis on originality, engagement, and viewer value.

Shorts at Industrial Scale

The stakes are high. According to Mohan, YouTube Shorts now average 200 billion views per day, making it one of the largest content surfaces on the internet.

Alongside AI avatars, YouTube is rolling out additional Shorts features, including image posts embedded directly into the feed and parental controls that allow guardians to limit, or entirely stop, Shorts scrolling for kids and teens.

These updates point to Shorts evolving from a creator tool into a tightly governed distribution system, where growth, safety, and monetisation intersect.

AI as Infrastructure, Not a Gimmick

Beyond avatars, YouTube continues to expand AI across its platform stack. Mohan highlighted upcoming tools that enable game creation through simple text prompts, as well as existing features such as AI-assisted editing, auto-dubbing in more than 20 languages, and AI-generated backgrounds for Shorts.

He also noted that over 20 million users have used YouTube’s Ask tool to gain deeper insights into the videos they watch, an example of AI being positioned as a contextual layer, not just a content generator.

Taken together, these initiatives suggest YouTube is treating AI less as a novelty and more as foundational infrastructure for creators, viewers, and advertisers alike.

Allowing creators to clone themselves for Shorts could change how content businesses operate, shifting value from time spent on camera to control over data, likeness, and distribution.

At the same time, YouTube’s emphasis on disclosure and quality controls indicates that scale alone will not be rewarded. In an era of AI abundance, platforms may increasingly compete on trust, governance, and audience experience.

For creators, the message is clear: AI may help you post more, but it won’t excuse posting less thoughtfully.