YouTube is rolling out new AI-powered features for Shorts and podcasts, from text-to-video generation with Veo 3 Fast to automated podcast clip creation.
Veo 3 Fast produces videos in 480p with low latency, and for the first time, it can generate audio alongside visuals. The initial rollout is happening in the US, UK, Canada, Australia, and New Zealand, with more regions to follow.
In Shorts, creators will gain new options: movements from videos can be transferred onto photos, letting a still image mimic a dance move. Style filters like Pop Art and Origami are included, along with the ability to insert props or characters through text descriptions. These updates will arrive over the next few months.
YouTube is also adding a "Speech to Song" remix feature, powered by Google's AI music model Lyria 2. It can transform spoken dialogue into songs, letting creators adjust moods such as "chill" or "danceable." The feature is set to launch soon in the US.
Another experimental tool called "Edit with AI" takes raw footage directly from a camera and generates a first cut. It automatically identifies key moments, adds transitions and music, and even delivers a real-time voice-over that reacts to what’s happening on screen. The tool supports both English and Hindi and is being tested for Shorts as well as in the YouTube Create app.
AI tools for podcasters
YouTube also announced new AI features for podcasters. In the coming months, US users will be able to generate automatically suggested clips from video podcasts, which can be instantly repurposed as Shorts.
For audio-only podcasts, YouTube plans to roll out an AI feature that automatically generates a customizable video for each episode. This capability will be tested in early 2026 with a small group of podcasters before expanding to a wider community.
