The creation of short videos is experiencing a new twist with the arrival of the AI-generated avatars on YouTube ShortsGoogle has begun rolling out a tool that allows creators to produce clips featuring a digital clone that looks and sounds like them, without the need to physically appear in front of the camera.
This new development comes at a time when the technology industry is trying to expand the use of Generative AI in video without further exacerbating the problem of deepfakes, scams and the flood of low-quality synthetic content, the so-called "slop" that already worries platforms, regulators and users.
What is the AI avatar for YouTube Shorts and how does it work?
The new feature, known internally as “My Avatar” or “Make a video with my avatar”It allows users to create a virtual double based on their own face and voice. The goal is to generate short videos for Shorts in which this digital clone appears speaking or gesturing following the instructions that the creator writes in a text.
YouTube describes these avatars as representations that “They will look and sound like you”, using in-house AI models, primarily Gemini and the Veo video technologyThe tool is designed to make the process relatively simple from a mobile device, taking advantage of the existing content creation options in the YouTube app and YouTube Create.
For now, avatar use is limited to clips up to eight seconds longAlthough several can be linked to chain scenes within the Shorts format, the platform leaves the door open for this technology to be extended to longer videos in the future, if it is well received, beyond the realm of shorts.
These “digital doubles” are especially useful for creators who want to maintain their on-camera presence but avoid showing oneself directly due to shyness, privacy or simple convenience. They can also be used to speed up the production of frequent content without having to record each take in a traditional way and, in some cases, export avatars to other applications.
Steps to create your avatar: live selfie and voice capture
Creating an avatar isn't a one-touch process, but YouTube has designed a fairly guided workflow. To begin, you need to meet two basic requirements: Be over 18 years old and have an active YouTube channelOnly then does the option to generate the digital clone appear within the application.
From the main YouTube app on your mobile device, the process involves tapping the button “Create” and select the Gemini icon in the upper right corner. Within the video creation tab, the section “Make a video with my avatar” or “Make a video with my avatar” is displayed, which is the entry point to the AI system.
At that point, the phase begins “Live Selfie”The platform asks users to record a selfie-style video in which they must pronounce several phrases indicated on the screen. This recording captures both the face and voice, and serves as the basis for the AI model to build the digital replica.
Google recommends recording with good lighting, clean background and no other peopleIn a quiet environment and with the mobile device at eye level, these conditions help the avatar better capture facial features, expressions, and voice nuances, reducing generation errors.
When the AI finishes processing the video, the app displays a avatar previewIf the result isn't entirely satisfactory, the user can repeat the process from scratch as many times as they like until they achieve a version they're happy with. Only when the creator clicks "Next" or the equivalent option will the avatar be saved and ready to use in Shorts.
How to use your avatar in new Shorts and in videos already published
Once generated and accepted, the avatar becomes available as the main character for the new YouTube Shorts that the creator wants to produce. From the Shorts creation menu appears the option “Create video. Make a video with my avatar”, which activates the specific AI interface.
In that environment, the user only has to Write a brief text description Gemini interprets the instruction and generates a clip in which the avatar speaks or moves according to the script. The initial duration is up to eight seconds per clip, with the possibility of linking several together to tell a more elaborate story.
In addition to creating content from scratch, YouTube allows Insert avatar into existing Shorts that are “eligible”. From the short videos section, the creator can click on “Remix” and then on “Reimagine”. The “Add me to this scene” option then appears, which integrates the digital clone into the selected video.
The company hasn't specified exactly what makes a Short eligible, leaving creators with some questions about when they can use their avatar to reinterpret previous content. In any case, the control rests with the channel owner, who decides whether their own videos can be remixed with AI.
For the moment, The scope of the function is limited to the Shorts format.The extension of these avatars to longer videos will depend on how the community responds and whether the tool fits into the close relationship dynamic that many viewers expect from creators.
Privacy, creator control, and avatar deletion
Beyond the creative impact, YouTube insists that the creator will maintain total control over your digital cloneAt any time, you can access the avatar management section to repeat the creation process, update it, or delete it if you change your mind.
If the avatar is deleted, the platform does not retain the materials used to train or generate that representation, according to information provided by the company. However, Previously published Shorts that feature the avatar are not automatically deletedThe creator will have to delete them one by one if they want to remove them from their channel as well.
Another relevant measure is the expiration policy: Avatars that remain inactive for three years will be automatically deletedThis rule introduces an unusual time limit in digital cloning tools, where normally the avatar remains linked to the account indefinitely unless otherwise requested.
YouTube also emphasizes that Only the creator can use their avatar.It is not possible for other users to use it to create videos, not even in collaborative mode. The idea is to prevent the digital clone from becoming an open door to impersonation within the platform itself.
These decisions come amid intense pressure on Big Tech to limit the misuse of generative AI. The company is thus trying to combine the appeal of a novel feature with a set of rules and safeguards designed to reduce risks associated with digital identity.
Watermarks, authenticity labels and the fight against deepfakes
All videos produced with these avatars incorporate visible and technical signs that indicate the use of AIIn practice, the generated Shorts will carry an on-screen mark to warn that the content is synthetic, along with special metadata that allows their origin to be traced.
Among those signs are labels such as SynthID and C2PAThese are standards that various platforms are beginning to adopt to document whether a video has been generated or modified using artificial intelligence. The idea is that both users and other services can more easily identify which videos did not originate from traditional recordings.
However, verification specialists point out that Labeling content is only part of the problemThe real challenge lies in ensuring that these watermarks are respected, not removed during re-uploads, and taken into account by those who share and consume the videos. Even so, YouTube considers these mechanisms a necessary step to avoid further confusion surrounding what is real and what is not.
The arrival of avatars comes as various reports, such as those from the Kapwing platform, warn of the growth of Low-quality AI-generated content flooding social mediaIt generates advertising revenue and displaces more elaborate materials. In this scenario, any tool that facilitates the creation of synthetic videos raises additional concerns.
Google maintains that its approach involves offering greater transparency and control tools to the creator, while simultaneously promoting its own generative AI strategy integrated into mass-market products like YouTube. The balance between innovation and trust will be key to preventing the feature from becoming a source of controversy.
Global rollout, requirements and the major European exception
The avatar tool is reaching users through a gradual deployment on a global scaleYouTube has not provided a detailed timeline or list of initial countries, but has clarified that the activation will be done in waves and that not all creators will see it available at the same time.
Access requires meeting three basic conditions: To be 18 years of age or older, have an existing channel, and use the official YouTube app or YouTube Create. on mobile. The company thus seeks to limit the use of the function to profiles with a certain track record and prevent, for now, the participation of minors in this form of digital cloning.
For those who meet the requirements in the included territories, the update should appear in the creation tab within the next few days. In fact, some users in different regions have already reported having the My Avatar system active, although it hasn't yet been rolled out to all accounts in some Asian markets.
The situation in Europe is different. For the moment, The feature will not be available in the European Union or in SpainSeveral sources indicate that regulatory hurdles and increasing demands regarding data protection and the use of generative AI are delaying the arrival of these tools in Europe.
This temporary exclusion fuels the feeling that European users are being left out of many AI innovations which can be tested in the United States, Latin America, or Asia. Faced with this situation, some choose to use VPNs or other workarounds, although doing so may violate the service's terms of use and does not eliminate the underlying legal concerns.
Google's bet on Gemini and Veo versus the closure of Sora
The launch of avatars for Shorts is not an isolated event: it is part of a larger initiative. Gemini's broader integration strategy within the YouTube ecosystem. The platform had already been testing features such as AI-powered automatic dubbing, generating clips from text, and intelligent assistants focused on channel analytics.
The combination of Gemini and Veo allows, according to the company, transform photos into video, create music and produce realistic imagesAll of this is done within the Google ecosystem. With this suite of tools, the tech giant aims to eliminate the need for creators to rely on external solutions to produce synthetic content.
In parallel, the move comes just as OpenAI has decided to shut down SoraIts experimental AI-powered video platform has been discontinued after months of struggling to make it a stable and socially accepted product. This setback has left a gap that other companies, including Google, are taking advantage of to strengthen their own offerings.
In this context, YouTube aspires to become the platform where AI-generated videos are created and distributed directlywithout the need for intermediate steps. For many users who previously created clips in external tools and then uploaded them to the channel, native integration can mean a time saving and a smoother experience.
The challenge, however, will be to prevent the platform from becoming filled with repetitive or purely synthetic content to the point of erode creativity and authenticity which have been their hallmark for years. The balance between productivity and quality will be crucial for Shorts's efforts to be seen as helpful and not just another source of noise.
With the arrival of AI avatars on YouTube Shorts, creators gain a powerful tool to appear in videos without being on camera, with clips of up to eight seconds controlled by text and clearly marked as generated by artificial intelligence; while the rollout progresses in phases and Europe remains, at least for now, outside the experiment, Google reinforces its commitment to Gemini and Veo in a field where creativity, identity and user trust will be more closely monitored than ever.
