New Delhi: Google has deployed the advanced AI video generator Veo 3 directly into YouTube shorts.
The integration, which began beta testing earlier this year, allows creators around the world to generate 8-second video clips from a simple text prompt with native audio within the familiar YouTube mobile app.
Powered by Google's VEO 3 model, this feature easily transforms imaginative ideas into sophisticated visuals. Users can describe scenes such as “Cat surfing in the waves of space” and “A bustling futuristic city at dusk,” and choose from styles such as movies and animation.
AI generates a standalone clip or green screen background that can be layered over shorts, allowing for remixing and personalization.
This is based on previous tools like Dream Screen, but it enhances them by enhancing the Veo 3's realism, speed and sound integration.
To access, the creator opens the YouTube app, taps Create, select Create from the Media Picker, and enters the prompt in English. The generated content is automatically watermarked with SynthID, a digital embedding that can be detected in each frame, and is labeled as AI-maids to promote transparency.
This development follows Google's first VEO announcement in May 2024, and VEO 2 made its shorts debut in February 2025.
YouTube emphasizes responsible use, urging creators to publish community guidelines and review AI output before complying.
The company is also expanding VEO 3 access via Ultra Plan with the Google AI Pro and Gemini apps. Users can upload photos to more customized generations.
As the shorts continue to dominate YouTube's daily 70 billion views, this VEO 3 injection could accelerate the virus trend, but risks an oversaturated feed with creepy AI artifacts.
