YouTube adds a Gemini editing assistant to Shorts

YouTube will add a Gemini-powered conversational editor to Shorts and its YouTube Create app, the company said at its Made on YouTube event on September 23, with the feature arriving in early 2027.

The assistant runs in a chat window. A creator asks for a first draft from a suggested prompt or writes their own, then keeps talking to it: reorder frames, trim pauses, sync music to the beat, add text overlays. At any point they can jump back into the manual timeline and adjust by hand, TechCrunch reported. YouTuber Happy Kelli, who demonstrated it on stage, described it as "a collaborative partner that edits with you, not for you."

YouTube's announcement says Gemini Omni's video intelligence went into Shorts and Create earlier in 2026, and that hundreds of thousands of channels used it per day in August. The company's first GenAI trends report found 72 percent of U.S. video posters aged 14 to 44 have used AI to help create or edit in the past year.

Most of the rest of the slate lands in YouTube Studio. Ask Studio, the Q&A feature, expands to the iOS and Android apps. A new feedback tool takes an unpublished draft and suggests changes to pacing, structure and storytelling. Creators can generate thumbnails and titles matched to their channel's style, and the dynamic thumbnail option takes three images and lets YouTube choose between them for different audience segments. Creators have run more than 40 million A/B tests on titles and thumbnails since the testing feature launched in 2024.

Next on that list is video A/B testing: up to three cuts of one video, fed to small audience segments and judged on watch time, with the creator picking the winner or YouTube doing it automatically after seven days, per The Verge. Editing a channel's back catalogue is the other half of the pitch: an agent watches older videos for ones that have become newly relevant, suggests new thumbnails or titles, and can assemble brand pitch material from a channel's own audience data, Amjad Hanif, YouTube's vice president of creator products, told The Verge.

The protections against synthetic video are being widened at the same time. Automated deepfake detection is available to channels to sign up for, and later this year YouTube will start combining speaking-voice detection with its facial likeness detection, then bring likeness detection into the mobile app.

Adobe has taken the other route to the same workflow, putting Firefly, Veo, Kling, Runway and Luma models inside the Premiere timeline rather than in a chat box.

If conversational editing works as shown, the assembly pass stops being the reason a Short publishes late, and the testable parts of a video's packaging spread further into the same panel. The Verge asked YouTube for data showing the existing AI tools improve metrics and the company declined to share any; Hanif's case for them is time saved.

Conversational editing arrives in Shorts and the YouTube Create app in early 2027; Studio's draft feedback and dynamic thumbnails are rolling out on web, iOS and Android, and live auto-dubbing is a pilot for early next year.