Netflix disclosed in its latest earnings that generative AI has appeared in roughly 300 of its productions so far in 2026. The revelation comes from the company’s Q2 shareholder letter and earnings call, where executives detailed how the technology now touches nearly every stage of filmmaking. From concept sketches to final visual effects, the streamer is betting big on these tools. But it insists humans remain firmly in charge.
The numbers surprised many in Hollywood. Engadget reported the figure directly from the letter, noting the heaviest concentration sits in post-production. Three titles received special mention. The Indian sports thriller Glory. The Brazilian soccer miniseries Brasil 70: A Saga do Tri. And the U.S. docuseries The American Experiment. Each used the tech to build complex sequences that would have been too expensive or time-consuming otherwise.
Ted Sarandos, Netflix co-CEO, elaborated during the call. He pointed to 17 minutes of AI-enhanced footage in The American Experiment. That material expanded the story’s scope in ways that simply wouldn’t have been feasible before. It wrapped up twice as fast and at half the cost of traditional methods. Sarandos didn’t stop there. “We believe it is going to enhance their abilities,” he said of creators, according to IndieWire.
His comments echoed across coverage. In Variety, Sarandos stressed that movies are still made by people who make movies. AI just supplies better tools. “I don’t think faster and cheaper matters if it’s not better,” he added. The message was clear. Quality comes first. Speed and savings follow only when the output meets a high bar.
This marks a sharp acceleration. Last year TechCrunch covered Netflix’s first use of generative AI for a final on-screen scene. It was a building collapse in the Argentine series El Eternauta. That sequence finished 10 times faster than with conventional visual effects. The 2026 update shows the experiment has scaled dramatically. Hundreds of titles. Broader applications. And growing confidence from the top.
Yet Netflix isn’t leaving the details to chance. The company published formal guidance for partners on its studio portal. Those rules lay out five guiding principles. Outputs must not infringe copyrights or recreate protected material. Tools cannot store, reuse or train on Netflix data. Work should happen in secure enterprise environments. Generated assets count as temporary aids, not final deliverables. And the technology must never replace talent performances or union-covered work without explicit consent. The full document sits at Netflix Partner Help.
A proposed use-case matrix helps teams decide when and how to proceed. It covers everything from mood boards in early ideation to digital alterations in visual effects. Some applications require extra review. Others get quicker approval. The goal is consistency. Transparency with Netflix contacts. And protection against unintended leaks or model contamination. Sarandos has said the company wants creators to feel empowered, not threatened.
Wall Street seemed to like the update. Netflix posted $12.56 billion in Q2 revenue, up 13.4 percent from a year earlier. Earnings beat expectations slightly. Investors have long pressed the streamer to control its massive content spend. AI offers one lever. If it trims budgets without hurting quality, margins improve. But the market also watches for audience reaction. So far backlash has been muted. Some viewers on social media noted odd visuals in recent releases. Others shrugged. The technology is already here.
Expansion continues. IGN detailed plans for 2026 that reach beyond production. AI will handle subtitle localization to reach more global viewers faster. It will generate custom ads for the ad-supported tier. Merchandising recommendations could grow smarter too. Sarandos told analysts the firm is “all in” on the technology. A later CNBC report captured that same phrase from the Q3 update, showing momentum building through the year.
Critics still raise familiar questions. Will AI displace jobs in post houses? Does it risk homogenizing visual style across shows? Netflix counters that its guidelines forbid training on proprietary data and require human oversight. In practice, many sequences start as rough AI drafts that artists refine. The building collapse in El Eternauta was one early test. The battle scenes and crowd enhancements in Glory represent the next wave. Each feels like a data point in a larger experiment.
Industry watchers note the competitive angle. Rivals are testing similar tools. But few have Netflix’s scale or its trove of viewing data to inform decisions. The company already uses machine learning for thumbnails, recommendations and encoding. Generative AI simply adds another layer. It can de-age actors, as in Adam Sandler’s Happy Gilmore 2. It can fill in backgrounds or simulate impossible camera moves. The creative palette widens. The cost curve bends.
Still, Sarandos returns to the same theme. “We’re confident that AI is going to help us and help our creative partners tell stories better, faster and in new ways,” he said, per the CNBC coverage. That order matters. Better comes before faster. The 300-title milestone proves the infrastructure works. Now the question shifts. How much further will Netflix push? And will audiences notice the difference?
Recent online chatter reflects the split. Some X posts celebrated efficiency gains. Others worried about a future where every explosion or epic vista looks suspiciously perfect. One thread linked back to the Variety piece and asked whether the “weird look” in certain Netflix originals finally made sense. The conversation is just beginning. Production teams, unions, regulators and viewers all have stakes.
For now the streamer shows no signs of slowing. Its internal AI studio acquisitions and new tooling suggest deeper integration ahead. The guidance document will likely evolve with experience. So will the use-case matrix. What started as a handful of experimental shots has become standard operating procedure for hundreds of shows and films. The genie is out. Netflix is learning how to direct it.
And the results so far? Faster workflows. Lower budgets on specific sequences. Expanded storytelling ambition in documentaries and scripted series alike. If Sarandos is right, the audience ultimately benefits. More ambitious visuals. Quicker releases. Stories that might never have been told. The technology itself stays invisible. The craft, he insists, remains human.
WebProNews is an iEntry Publication