YouTube rolled out a suite of generative artificial intelligence features on September 23, 2026, during its annual Made on YouTube event, integrating multimodal tools across video editing, recommendation feeds, and channel analytics. The platform introduced conversational editing powered by Google’s Gemini Omni model, allowing creators to issue voice or text prompts to trim footage, build custom viewer feeds, and manage content optimization.
Conversational Editing Powered by Gemini Omni
The platform integrated Google’s Gemini Omni model to let creators edit videos through natural language prompts rather than traditional timeline interfaces. According to YouTube, a creator can input commands such as removing empty silent sections or starting a video at its most engaging highlight, prompting the system to select best cuts, organize timelines, apply transitions, and generate automatic captions.
YouTube CEO Neal Mohan stated during the event that the platform has deployed over 3,000 product updates throughout 2026 to match shifting viewer habits and creator demands. Mohan emphasized that the new artificial intelligence utilities function as assistants to reduce production friction rather than substitutes for human creativity.
Custom Feeds and Viewer Directed Discovery
Viewers gain the ability to generate tailored recommendation algorithms through natural language prompts via the new Custom Feeds feature, which rolls out initially to users in the United States. Users can describe a specific mood or topic, such as requesting relaxing cooking videos for a particular frame of mind, and the system constructs a dedicated, separate video tab distinct from the primary homepage algorithm.
Alongside discovery changes, YouTube upgraded its backend analytics hub, YouTube Studio. The update introduces video A/B testing that permits creators to upload up to three distinct edits of a single video and compare audience retention rates to determine the strongest cut. Dynamic thumbnails automatically display different images to distinct viewer groups depending on expected engagement, while an analytics chatbot named Ask Studio reviews entire channel catalogs to suggest title and thumbnail optimizations.
Security Upgrades and Expanded Ecosystem Tools
To address synthetic media risks, YouTube expanded its likeness protection systems beyond facial recognition to identify and report unauthorized reproductions of unique human voices. The platform also updated its shopping tools with an interactive version of Ask YouTube, enabling viewers to request detailed product comparison tables and specifications directly while watching content.
Additional platform updates include an auto-dubbing translation feature for live streams starting in English and Spanish, a Shorts Series feature to group episodic short-form videos, and Ask Music within YouTube Music for natural language playlist generation. The ecosystem also added a podcast AI summary feature called Your Podcast Lineup, alongside live-stream donation competition tools known as Live Showdowns and voice comments designed for television viewers.