Netflix says 300 titles used generative AI this year — and that it's just getting started
In its Q2 2026 results, Netflix revealed that around 300 titles used generative AI this year, mostly in post-production, with revenue of 12.56 billion dollars, up 13% year on year.
Netflix has put numbers on the debate Hollywood has been having for two years: around 300 of the platform’s titles used generative AI this year, mostly in post-production. The figure landed in Thursday’s second-quarter results, alongside a phrase that says everything about intent — usage is “scaling quickly”.
Where is Netflix actually using generative AI?
Mostly behind the camera: effects, clean-up and image finishing. Co-CEO Ted Sarandos offered a concrete example: the documentary series The American Experiment carries 17 minutes of AI-enhanced footage, produced twice as fast and at half the cost of traditional methods. It is the economic argument in its bluntest form — and the reason Hollywood’s unions read these numbers with one eyebrow raised.
On the money side, the quarter was solid without dazzling: 12.56 billion dollars in revenue, up 13% year on year and a shade under Wall Street’s bar, with a third-quarter forecast that also came in shy of expectations. For the full year the company now points to 51 to 51.4 billion in revenue at a 31.5% operating margin — the official documents are on its investor relations site.
Will viewers notice the AI?
Increasingly yes, even if invisibly — the idea is for AI to speed up what was slow and expensive, not to write the scripts. But the line is slippery, as Netflix knows better than most: the same week it disclosed the 300 titles, it was digesting the flop of In the Hand of Dante, proof that stars and budget don’t replace a story. And audiences have already shown synthetic spectacle has limits — see the reception of the AI-generated Odyssey knock-off.
Three hundred titles in, the question is no longer whether AI enters the editing room. It is whether anyone notices — and Netflix is betting millions that the answer is no.
By Lucy Bennett
Image: Coolcaesar / Wikimedia Commons (CC BY-SA 4.0)
