Browse
Log in / Register
Share
Updated May 5, 2026
A top Netflix production executive said artificial intelligence has been woven into about 300 titles, but the company still draws a hard line between automation and emotion. Sung Q Lee, Netflix's senior director of production for Asia-Pacific, made the remarks at the Busan International Film Festival, where the streamer hosted its annual Creative Asia event with the Motion Picture Association. According to The Hollywood Reporter, Lee described AI as a production assistant, not a creative replacement.
Lee said AI is used across multiple stages, including previsualization, production design, set-dressing concepts, editing and visual effects. He framed the company's approach as a practical one: instead of asking how much AI can be packed into a title, he said the useful question is what a title can achieve with AI help. That framing positions AI as a resource rather than a spectacle, and suggests Netflix is embedding the technology into routine workflows instead of reserving it for tech-driven stunts.
The panel was co-hosted by the Motion Picture Association and featured Korean filmmaker Kang Yun-sung, known for "The Outlaws," and Kimball Thurston, chief technology officer at visual effects company Weta FX. The discussion repeatedly returned to two themes: AI's ability to stretch modest budgets, and the need to safeguard human creative judgment. Those two ideas are not always in easy balance, and the differing reactions in Hollywood and Asia highlight the tension.
According to The Hollywood Reporter, the Asian content sector largely sees AI as an opportunity, while parts of the Hollywood creative community continue to treat it with suspicion. That split is not just cultural. It also reflects different production cost structures and different levels of exposure to AI-driven labor disputes. In Asia, where budgets for many series and films are leaner, generative AI can be the difference between a project getting made and being shelved. In the U.S., the same technology is often discussed through the lens of contracts, residuals and job security.
Lee acknowledged the emotional side of filmmaking remains stubbornly human. New tools have helped with how films are made, he said, but they have never decided why a story is told. He added that the human touch is critical when a film needs to convey emotion. That is not a small caveat; it speaks to a growing concern among critics that AI-assisted productions may flatten tone and performance.
Kang Yun-sung has moved from curiosity to production experience with AI. Last year he co-directed "Run to the West" with Kwon Han-seul, a film billed as Korea's first fully AI-assisted feature. It earned about $100,000 at the box office and drew criticism for its AI-generated visuals. Kang acknowledged that the technology advanced so quickly that the film already looked somewhat dated by release. That might sound like an argument against AI, but Kang says the experience was evidence of AI's potential rather than proof of its limits.
His next project, the action film "The Gulf of Aden," leans heavily on generative AI tools from Kling AI, a Chinese company. Kang told the panel the tools allowed him to execute scenes that budget constraints would otherwise have ruled out. He estimated the approach would lower the film's production costs by about 40%. This is not only a matter of replacing digital effects with cheaper alternatives. Kang said the live-action shoot itself was streamlined, which reduced production time.
That level of cost reduction is likely to draw attention from U.S. producers wrestling with rising production expenses. Even if the quality bar remains controversial, the efficiency argument is becoming harder to dismiss.
Weta FX CTO Kimball Thurston described his company's use of AI as "assistive AI." In his view, the technology should not be allowed to alter creative intent. Instead, AI moves quickly to produce visuals that directors and cinematographers can then review, spot problems in, and correct. Thurston said the goal is to reduce simple, repetitive tasks, which frees artists and technicians to focus on more demanding creative problems.
The phrase "assistive AI" is useful for the broader industry. It draws a line between tools that support a human's vision and systems that autonomously generate decisions. Netflix's description of its own AI work, at least as presented in Busan, falls on the same side of that line. Lee and Thurston both portrayed AI as a scalable helper, not a signature artist.
Netflix's decision to quantify its AI usage is a notable benchmark. The streaming giant did not disclose a title-by-title breakdown, so it is not possible to measure how heavily each of the 300 productions relied on the technology. But the scale itself signals that AI is no longer an experimental add-on in Netflix's production ecosystem. It is becoming part of the studio's standard toolkit.
For U.S. producers, that raises questions about disclosure, talent agreements and creative credit. If a platform like Netflix can apply AI across hundreds of titles without making it the centerpiece of its marketing, the technology may be entering the mainstream in ways that are easy for audiences not to notice. The Busan panel was framed around creative benefits, but the broader U.S. debate is likely to focus on transparency: which parts of a production were touched by AI, and who should be credited when those touches shape the final story.
There is also a practical layer. If Kang's 40% cost reduction estimate proves repeatable on larger English-language productions, AI-assisted filmmaking could quickly move from a niche experiment to a standard line item in production budgets. That would put significant pressure on projects that do not adopt some form of AI workflow.
The writers and actors strikes in recent years underscored how sensitive U.S. talent is to technology changes. The message from Busan, however, was not about eliminating jobs. It was about shifting labor from repetitive tasks to creative judgment. That may not fully resolve labor anxieties, but it does put forward a framework in which humans remain in charge of why a story is told, while AI handles more of how it is made.
Lee's comments, combined with Kang's production data and Thurston's cautious tech-industry view, paint a consistent picture. AI in filmmaking is spreading quickly and quietly. It is being used for concrete tasks, and it is being deployed to save money. But the emotional core of a film, the part that makes an audience care, remains tied to human direction. The technology can assist, optimize and accelerate; it cannot yet supply the instinct that decides which moments matter.
For American readers, the practical takeaway is simple: AI is already inside the standard production process at the world's largest streaming platform, and more titles will follow. The open question is not whether AI will be used, but how clearly studios will explain that use to audiences and to the talent who help make the work.
