Ted Sarandos, co-chief executive of Netflix, told analysts on the company's second-quarter earnings call on 16 July that generative AI had been used across roughly 300 Netflix titles in 20261. Generative AI here means software that produces new footage, audio or images rather than editing existing material. Sarandos said the use sits mostly in post-production, the editing, visual-effects and finishing stage after shooting ends, and is now moving forward into concept work and previsualisation.
He named the stack for the first time: Interpositive, the production company founded by Ben Affleck and bought by Netflix in March for up to $600m; Eyeline, an internal tool; and Animation Lab, Netflix's in-house animation capability. Three productions came with it, Glory, Brasil 70: A Saga do Tri, and The American Experiment, the last carrying around 17 minutes of AI-enhanced footage that Sarandos said was made twice as fast and at half the cost of the conventional route.
Sarandos added that the money saved goes back into content on the service rather than out as margin. For anyone selling tooling into the studio segment, that is the sentence worth quoting in a pitch. At Netflix, AI cost reduction is being spent as commissioning budget, which means the buying decision sits with content leadership rather than with a chief financial officer hunting savings.
The disclosure closes a question this beat opened in May. When Netflix confirmed Netflix INKubator, its generative-AI animation unit under Serrena Iyer, on 15 May, it declined to name the vendor or model underneath the pipeline. Buy and build now carry three names, a purchase price and a title count.
