Netflix Adds an AI Tab to Help You Pick What to Watch

Netflix is testing a new way to browse its catalog: instead of scrolling endless genre rows, you'll be able to tap a themed category like "Something Light," "Family Movie Night," or "Quick Watches" and get a curated set of recommendations built around that mood. It's called AI Discover, and it's rolling out as its own tab in the top menu of Netflix's TV app, according to The Verge.

The mechanics are simple on the surface. Pick a theme, get suggestions. Don't like what shows up, tell Netflix, and the app remembers that feedback for next time. Netflix describes the system as running on "a new intelligence layer we've been building that responds to what members tell us in real time and keeps learning," per the company's statement to The Verge.

That's a notable shift from how Netflix recommendations have traditionally worked. The service has long leaned on viewing history and algorithmic pattern-matching to guess what you might want to watch next, sorting things into familiar buckets like genre or "because you watched X." AI Discover instead organizes content around a feeling or occasion, the kind of framing you'd normally do yourself before asking a friend for a recommendation. Running it through a generative AI-style system rather than a static genre taxonomy also means the categories themselves can, in theory, evolve rather than stay fixed.

The rollout is limited for now. AI Discover is starting gradually in the US, limited to people using Netflix in English on supported TVs, so this isn't a flip-the-switch global launch yet. Netflix is explicitly framing it as an experiment, which tracks with how the company has approached other AI features: test with a slice of users, see what sticks, expand or scrap accordingly.

Why this is landing now

The timing is hard to ignore. The announcement comes shortly after Netflix laid off roughly 1,000 employees, a move tied to slowing growth and reports that users are spending less time on the service than before. A feature explicitly designed to help people actually find something to watch is a reasonable response to that pressure. Discovery has been one of streaming's persistent pain points: with thousands of titles across a service, "I don't know what to watch" is a real reason people open an app, scroll for a few minutes, and close it without pressing play. If AI Discover meaningfully cuts down that friction, it's a direct lever on engagement, and engagement is the metric Netflix most needs to defend right now.

It also fits into the broader AI arms race playing out across every major consumer platform. Netflix hasn't been as loud about generative AI as, say, the big search or productivity companies, but a recommendation engine that responds to real-time feedback and "keeps learning" is Netflix's version of the same pitch everyone else is making: let AI handle the friction, let the product feel more responsive to you personally.

What's still unclear is how much of this is genuinely new technology versus a repackaging of existing recommendation infrastructure with a more conversational front end. Netflix hasn't detailed the underlying models or how AI Discover differs technically from its existing algorithm beyond the themed framing and the feedback loop. For a company that's historically kept its recommendation system closely guarded, the vague "intelligence layer" language leaves plenty of room for interpretation.

For now, the practical takeaway for US subscribers on supported TVs: watch for a new "Discover" tab, try a theme, and tell Netflix when it misses. Whether this stays a modest convenience feature or becomes the new default way people browse the service will depend on whether Netflix can show the data behind it, something the company tends to reveal only once a feature has proven itself at scale.