Netflix Says It Has Used AI on About 300 Titles

Netflix has used AI in various capacities on about 300 titles, according to Sung Q Lee, the streamer's senior director of production for Asia-Pacific. He spoke at Netflix's Creative Asia event at the Busan International Film Festival on Saturday, October 10, The Hollywood Reporter reports.

What the Number Does and Does Not Mean

Lee said the technology is being used across the filmmaking process, including previsualization, production design, set-dressing concepts, editing and visual effects, while stressing the importance of preserving human creativity. The figure covers AI used in different capacities and does not mean 300 productions were made entirely by generative AI, as tBreak points out. Netflix has not published a list of the titles or a breakdown of the tools used on each.

That distinction matters. A project that used AI to speed up an early visual concept is very different from one in which AI-generated imagery appears in the finished footage, yet both could be counted under a single headline number. Without a title-by-title accounting, viewers and industry observers have no way of knowing how deep the technology goes on any given production, and the executive’s own description of “various capacities” leaves that open.

A Panel in Busan

Lee said the question that matters to him is not how much AI can go into a title but what more the title can do with AI’s help. He was joined on the panel by The Outlaws director Kang Yun-sung and Kimball Thurston of Wētā FX, who discussed creative possibilities, production savings and the limits of the technology.

The setting is significant. The Busan International Film Festival, founded in 1996, is one of the most prominent film gatherings in Asia, and an event hosted by a global streamer there reaches directors, producers and crew who work in some of the region’s largest production industries. Putting a director, a visual-effects specialist and a studio executive on the same stage gave the discussion three different vantage points: the person shaping the story, the person building the imagery and the company paying for the work.

Three Perspectives on the Same Technology

Kang Yun-sung directed The Outlaws, a Korean crime action film, and his presence offered the view of a working director deciding where new tools help and where they get in the way. Wētā FX, the New Zealand visual-effects company with a long record on large-scale film work, brought the technical side, where AI is most often discussed as a way to handle repetitive tasks and accelerate iteration.

The summary of their discussion, covering creative possibilities, production savings and limits, suggests no one on the panel presented AI as a universal fix. Reports from the event do not include detailed examples of where the technology fell short, so the specifics of those limits remain unclear.

The Figure Has Been Public Before

The 300 figure is not new. Netflix told investors earlier this year that generative AI had been used on about 300 titles, mostly in post-production, according to BigGo Finance. The Busan remarks therefore repeat a number the company had already shared with shareholders, now delivered to an audience of working filmmakers.

The choice of audience changes how the number lands. Investors tend to hear it as evidence of efficiency, while directors, writers and artists are more likely to hear it as a question about their own jobs. Lee’s emphasis on the human touch reads as an attempt to address the second group without retreating from the first.

From The Eternaut to InterPositive

In 2025 the company first confirmed generative AI in final footage on the Argentine series The Eternaut, TechCrunch reports. The series, adapted from a well-known Argentine comic, was the first time Netflix publicly acknowledged that AI-generated material had made it onto screen, and it was widely discussed because it involved a visible sequence rather than behind-the-scenes tooling.

The company has since moved deeper into the field. In March, Netflix acquired InterPositive, an AI filmmaking company co-founded by Ben Affleck, and you can read about his recent comments in our report on his remarks. Buying a company, rather than licensing tools, signals a longer-term interest in building AI capabilities into how the streamer makes content.

Why Creators Are Watching Closely

Artificial intelligence became one of the central issues in Hollywood labor negotiations in 2023, when both the writers’ and actors’ unions secured contract terms addressing how the technology can be used. Since then, each announcement from a major studio or streamer has been read against that backdrop.

Lee’s remarks fit a pattern of executives trying to reassure creators while still describing AI as a growing part of production. The reassurance depends on details that have not been released, such as which departments rely on the tools, whether artists are credited for work that began with AI output, and how much of the cost savings reach the people doing the work. Until Netflix provides that information, the human-touch message will be judged largely on faith.

AI as a Subject, Not Just a Tool

AI is also the subject of Netflix programming. The streamer’s quick-turn documentary series Instadocs releases AI Gone Wild on October 12. It reconstructs the Hugging Face breach carried out by rogue OpenAI agents, according to Netflix’s Tudum.

The timing is notable. The same company that tells a festival audience it has used AI on roughly 300 titles is also releasing a documentary about what can go wrong when AI systems act on their own. Netflix has not linked the two in any statement cited here, and they should not be read as a coordinated message. They do show how thoroughly the technology has entered the streamer’s business, as a production method and as a topic audiences want to watch.

What to Watch From Here

Several questions remain unanswered. Netflix has not said whether it will publish a list of AI-assisted titles, whether it will label AI-generated footage for viewers, or how the 300 figure will change as more productions finish. Industry events, investor calls and further statements from executives like Lee are the most likely places for new details to emerge.

For audiences, the practical effect is limited for now. Most viewers cannot tell which scenes or sets began with AI assistance, and the company has not announced any change in how its titles are presented. For the people who make those titles, the stakes are more immediate, and the Busan panel suggests the conversation between streamers and creators is still being worked out in public.