Netflix Details How AI Is Reshaping Film Production

Netflix says artificial intelligence has now been used in some form across roughly 300 of its titles, offering one of the clearest snapshots yet of how the technology is entering mainstream film and television production. The figure was shared during a Creative Asia panel at the Busan International Film Festival, where speakers emphasized assistance and efficiency rather than the replacement of human storytellers.

Where Netflix says AI is being used

Lee Sung-gyun, Netflix’s senior director of APAC production, described applications in previsualization, production design, set-dressing concepts, editorial work and visual effects. One presentation showed hybrid production methods combining live-action footage, VFX and generative tools, including crowd scenes that would have been difficult to deliver under tighter schedules and budgets.

The “300 titles” figure covers varied uses, not 300 fully AI-generated productions. That distinction is central to the company’s argument. Lee said Netflix tests tools under internal guidelines and does not adopt them simply to be first, while maintaining that human judgment remains necessary for emotion and story.

A specific cost claim, not an industry average

South Korean filmmaker Kang Yoon-sung said AI workflows could reduce a feature budget by 30% to 40%, citing planning for his maritime action project The Gulf of Aden. He attributed the projected savings not only to fewer conventional CG and VFX demands, but also to a simpler shoot and a shorter production period. The estimate relates to that project and should not be read as a guaranteed saving for every film.

Wētā FX chief technology officer Kimball Thurston also outlined the limits. He said the studio has introduced machine-learning tools at a measured pace, including image cleanup and motion for crowd scenes, while flagging continuing problems with color accuracy and the lack of interfaces designed around visual artists rather than text prompts.

All three speakers resisted the idea that actors are the obvious target for replacement. Their shared position was that AI can expand the number of creative options, make difficult shots affordable and improve repetitive workflows, but cannot supply the lived emotion at the center of performance.

The Busan discussion therefore presented a more complicated picture than either hype or panic: AI is already inside major production pipelines, but its most credible near-term role may be as a controlled technical layer shaped by filmmakers, artists and strict production rules.