Netflix used generative AI on about 300 of its titles this year, the company told investors alongside its second-quarter results on July 16. The figure covers productions that used the tools somewhere in their pipeline, with “the largest concentration of work in post-production,” according to the shareholder letter — a number that landed as one of the most-discussed lines in the whole report, as trade coverage noted.
The 300 figure needs one clarification before anything else: it counts titles that touched generative AI at some stage, not 300 films conjured out of nothing. The work ranged across concept art, pre-visualisation, visual-effects references and post-production, and it is the last of those where most of it happened.
What the generative AI actually did
Co-CEO Ted Sarandos gave a concrete example: a US docuseries, The American Experiment, that used the tools to build sequences a documentary budget could not normally reach.
“It enabled us to expand the scope of the series in ways that wouldn’t have been feasible before. Those 17 minutes, they were produced twice as fast and at half the cost of previous options.”
Seventeen minutes of a single series is a lot of screen time to attribute to a tool that barely existed at production scale a few years ago. Sarandos named other titles too — the Indian series Glory and the Brazilian soccer miniseries Brasil 70 — in describing enhanced crowd sizes, historical battle scenes and worldbuilding shots.
The argument Netflix is making
Netflix’s case is not that AI replaces filmmakers but that it rescues shots they would otherwise abandon. Sarandos put it directly.
“Keep in mind that in many of the cases, productions would’ve left out those key shots because they wouldn’t have been able to afford them, they wouldn’t have been able to do them in the time frames that they’re working on. So those sequences are saved by the availability and access to the Gen-AI tools.”
That is the optimistic framing, and it has a business end. “By equipping creatives with these tools, we believe it is going to enhance their abilities and have better and more impact for every dollar we spend on our programming,” he said. The savings, the company says, get reinvested into more content — the same profit-over-revenue logic driving its quarterly financial results.
What Netflix did not dwell on
The framing is entirely Netflix’s, and it skips the part that made the 2023 Hollywood strikes about exactly this. The Writers Guild and SAG-AFTRA walked out in large part over AI, and their contracts now govern consent, credit and reuse. Netflix’s announcement did not address how the 300 titles handled those terms, and no union issued a specific response to this disclosure — which is its own kind of silence.
There is a craft worry underneath the labour one, too: that heavy, cost-driven use of the same tools produces a sameness, a homogenised look creeping across a service’s output. Netflix has already leaned on AI for more overt tricks, including using it to recreate a late actor’s voice, which is where the consent questions get sharpest.
Why the number keeps climbing
Netflix’s appetite for these tools is not incidental. The company frames generative AI as a lever on the single thing it now asks investors to judge it by — profit margin — and a tool that makes 17 minutes of footage at half the cost is, to a studio, a margin story before it is an artistic one.
That is what makes 300 a floor rather than a ceiling. If the tools save money and time on the sequences productions used to cut, the incentive runs one way. Next year’s number will almost certainly be larger, and the questions Netflix answered lightly this time — about who consents, who gets credited, and what all of it does to how the output looks — will get harder to wave past.
For now, the disclosure functions as a marker. A studio that once treated AI as a quiet backroom tool is now citing it by the hundred to investors as evidence of discipline. That shift, from something to downplay to something to advertise, may be the real news in the number — a sign of how quickly generative AI has moved from experiment to expectation inside the business of making television.
