Loading page…
AI Discover lets viewers narrow suggestions by mood, audience or available time—and reject titles they do not want recommended again. The experiment excludes kids’ profiles.
Listen to this story
The test turns recommendation discovery into an explicit, iterative flow: subscribers can select preset themes, narrow matches, or reject a title, while Netflix still factors in viewing history, preferences, and platform trends. That gives members a way to express what they want now rather than rely only on inferred taste, but Netflix has not demonstrated better recommendations; it describes the underlying system as a new intelligence layer. Access is limited to some English-language U.S. accounts on supported TVs, with expansion planned in the coming weeks.
The rollout began November 3, 2026, and excludes kids’ profiles.
Themes include moods such as “Something Light,” audience choices such as “Family Movie Night,” and options based on available time or trending titles.
A “Tune Your Recommendations” control lets viewers keep a suggestion or tell Netflix not to show that title again in AI Discover.
Netflix is giving some TV viewers a more direct way to steer what it recommends. On November 3, 2026, it began rolling out AI Discover to a limited group of U.S. subscribers using English on supported smart TVs. The experiment starts with choices about mood, audience and available time—not a conversation with a chatbot.
Netflix has long used machine learning and AI to personalize recommendations. What changes here is the presentation and the interaction: Variety describes AI Discover as the company’s first user-facing application branded as AI-powered. Viewers enter a dedicated discovery area and make explicit selections, rather than only receiving suggestions informed by their viewing history and preferences.
Subscribers included in the rollout will see a Discover tab in the TV app’s top navigation bar. It carries a red gem-shaped icon and a New label. AI Discover is unavailable on kids’ profiles.
The first experience inside the tab uses preset themes, each previewed with a clip from a film or show. As TechCrunch details, those choices cover several kinds of intent: what viewers feel like watching, their preferences, how much time they have, and what is new or trending.
Choosing a mood does not replace Netflix’s existing knowledge of a viewer. Netflix says the recommendations are produced in real time using viewing history and preferences, alongside what is trending on the platform. The feature therefore combines a member’s immediate choice with signals about their tastes and the wider service.
The company describes the underlying technology as a new intelligence layer that responds to user input in real time and keeps learning. That is Netflix’s account of how the system works, not a measured improvement in recommendation quality.
AI Discover is also a testing space, not a settled redesign of Netflix. The company says customer use and feedback will help determine which experiments it keeps or improves, and which might enter the main Netflix experience. It plans to extend access to more customers and regions in the coming weeks.
Loading discussion...
Join the conversation
Explain when past tastes help—and when you want a fresh start.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.