YouTube Launches AI-Powered Custom Feeds for Personalized Content

YouTube is rolling out a feature that gives users direct control over their content recommendations through AI-generated custom feeds, the company announced Wednesday at its Made on YouTube event.

The new capability allows users to type natural language descriptions of the video content they want to see, with Google's Gemini AI model building personalized feeds based on those requests. These custom feeds appear as pinned tabs at the top of the YouTube home page, separate from the platform's main recommendation feed.

How custom feeds work

Users create feeds by entering detailed prompts describing their preferences. YouTube's examples include requesting "video podcasts to watch during your 30-minute train commute" or "relaxing commentary videos to unwind with." The AI processes these descriptions to curate relevant content from YouTube's library of over 20 billion videos.

The prompt-based system accepts lengthy, detailed instructions. Users can specify what content to include or exclude, which topics to prioritize, and other curation parameters. Multiple custom feeds can be created and accessed alongside YouTube's standard recommendation feed.

Why it matters

Custom feeds represent YouTube's response to a growing trend across social platforms toward user-controlled algorithms. As recommendation systems face increasing scrutiny over filter bubbles and engagement optimization, giving users explicit control over their content diet addresses both transparency concerns and diverse use cases. For YouTube specifically, the feature could help users navigate its massive content library more effectively while reducing friction between the platform's algorithmic goals and individual user preferences.

Following industry momentum

YouTube joins several other platforms that have recently introduced similar algorithmic customization tools. Bluesky pioneered customizable feeds as a core differentiator, later adding an AI tool called Attie to simplify feed creation. Meta's Threads and Instagram, X (formerly Twitter), and Spotify have all launched their own versions of user-controlled feed-building features, many incorporating AI assistance.

Emily Moxley, YouTube's VP of Product Management for Viewer AI, framed the feature as an exploration tool. "One of the amazing things about being on YouTube is that the world is always at your fingertips," Moxley told reporters. "There's over 20 billion videos in the YouTube corpus, so you're always one search away from a treasure trove of topics."

The custom feeds feature will begin rolling out next month across YouTube's web and mobile platforms, according to the company. Details were first reported by TechCrunch.