YouTube Shorts: AI Avatars for Secure Deepfakes
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
YouTube Shorts Enriches with AI Avatars
YouTube Shorts, Google's short video platform, has recently introduced a new AI-based feature that allows content creators to generate realistic digital avatars of themselves. This initiative, announced earlier this year, highlights the complexity of YouTube's relationship with AI-generated content, adding generative tools while seeking to limit abuses related to deepfakes and scams.
Users can now create an avatar, a digital version of themselves, which can be inserted into existing videos or used to create new ones. According to YouTube, these avatars will replicate the user's appearance and voice, providing a safer way to use AI for content generation.
A Structured Creation Process
To create an avatar, users must record a "live selfie" that captures their face and voice while following specific instructions. YouTube recommends ensuring good lighting, a quiet environment, and holding the phone at eye level to achieve the best results.
Once the avatar is created, users can select the "make a video with my avatar" option to generate a clip from prompts, with a maximum duration of eight seconds. Avatars can also be added to "eligible Shorts," although YouTube has not specified the eligibility criteria.
Restrictions and Enhanced Security
The use of AI avatars is subject to strict restrictions. They can only be integrated into the creator's original videos, who retains control over the ability to remix their Shorts. Creators can delete their avatar or associated videos at any time. Additionally, avatars that remain inactive for three years will be automatically deleted.
All videos using avatars will be clearly identified as AI-generated, thanks to a visible watermark and digital labels such as SynthID and C2PA.
Gradual Rollout and Access Conditions
The feature will not be immediately available to all users. YouTube plans a gradual rollout, without specifying a timeline or geographical launch areas. Creators must be at least 18 years old and have an existing YouTube channel to access this tool.
This new feature adds to YouTube's growing range of AI tools, which already includes AI-generated video clips, automatic AI dubbing, and a channel analysis chatbot. These innovations are largely powered by Google's Gemini AI models.
Competitive Context
The launch of this feature coincides with OpenAI's withdrawal from the video generation market. The startup recently announced the end of its video tool Sora after facing challenges related to copyright and content quality, making the project unattractive to investors in light of a potential IPO.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.