Netflix Acquires Ben Affleck’s InterPositive AI Filmmaking Company

Industry NewsAI

by Johnnie BehiriMarch 9^th, 2026

Subscribe to News - Industry

Less than a week after walking away from its proposed $82.7 billion acquisition of Warner Bros. Discovery, Netflix has made a very different kind of deal: the full acquisition of InterPositive, an AI filmmaking technology company founded in 2022 by Ben Affleck. The entire 16-person team of engineers, researchers, and creatives will join Netflix, with Affleck signing on as Senior Advisor. Financial terms were not disclosed.

The timing is pointed. Netflix spent months pursuing the largest media consolidation deal in history, faced a DOJ second request and bipartisan political opposition, and ultimately declined to counter Paramount Skydance’s improved bid, with co-CEO Ted Sarandos stating the transaction was “no longer financially attractive.” Days later, the company turned around and announced a targeted, creator-led acquisition of a company most of the industry had never heard of. InterPositive had been operating entirely in stealth since its founding, incorporated under the name Fin Bone LLC, and had never been publicly disclosed until Netflix’s announcement.

What InterPositive actually built

InterPositive’s technology is not what most filmmakers might initially assume when they hear the words “AI” and “Netflix” in the same sentence. It is not a text-to-video generator, and it does not synthesize performances. Affleck has been explicit about this distinction, describing the category gap clearly in the video Netflix released alongside the announcement: existing generative AI models, he argues, are focused entirely on the final visual output, without any understanding of how cinematographers and directors actually construct shots, manage production variables, or make creative decisions under pressure.

What Affleck and his team built instead is a workflow that starts on set and stays there. InterPositive trains a custom AI model on a production’s own dailies, using that footage as the foundational dataset rather than pulling from public internet sources. The resulting model learns the specific visual language of a particular project: its lighting signatures, its lens characteristics, its established editorial logic. Filmmakers can then use that model in post to solve concrete production problems: removing stunt wires, relighting shots, reframing to correct framing errors, recovering shots that were missed, or enhancing backgrounds without resorting to traditional VFX pipelines.

The soundstage dataset and the deliberate constraints

To build the initial model, Affleck and a small R&D team filmed a proprietary training dataset on a controlled soundstage designed to replicate the conditions of a full production. The goal was to give the model fluency in the vocabulary cinematographers and directors already speak, while exposing it to the kinds of inconsistencies and failures that are endemic to real shoots: missing coverage, incorrect lighting, and continuity issues between takes. That specificity was the point.

Equally significant is what Affleck deliberately chose not to include. The models are focused on filmmaking techniques rather than performances. That is not a minor technical detail. It is a boundary the company built in at the architecture level, ensuring the system cannot be used to manipulate, replace, or synthesize human performances, which remains the sharpest fault line in Hollywood’s ongoing AI debate. “We also built in restraints to protect creative intent,” Affleck wrote in Netflix’s announcement, “so the tools are designed for responsible exploration while keeping creative decisions in the hands of artists.”

This positions InterPositive at some distance from the broader generative AI landscape that has made the creative community deeply anxious. As we covered in our Netflix Generative AI Guidelines analysis, the streamer has already drawn clear internal lines around talent consent and performance synthesis. The InterPositive architecture appears to have been designed with those same concerns as foundational constraints rather than afterthoughts.

Netflix’s AI trajectory and what this acquisition signals

Netflix has been building toward this kind of acquisition for years, even if the specific move was not predictable. We previously covered Netflix’s first public use of generative AI for VFX footage in the series “El Eternauta,” where the company claimed a 10x efficiency improvement over traditional methods. We also reported on DifFRelight, the facial relighting framework from Netflix Eyeline Studios, which demonstrated the kind of diffusion-based post-production control that aligns closely with what InterPositive is building toward. The pattern has been consistent: Netflix develops or acquires production-side AI tools aimed at reducing cost and friction in post, while publicly insisting that creative judgment remains human.

What is different about InterPositive is the model’s origin. Until now, Netflix’s AI tools have been built internally by engineering teams. InterPositive was built by a filmmaker and his collaborators, born of frustration with what existing models could not do. Affleck said in 2022 he began observing the early rise of AI in production and found that the models available “came up short” for filmmaking purposes, with no fluency in cinematic logic or tolerance for real-world production constraints.

Netflix has confirmed it does not plan to sell access to InterPositive’s tools externally. That last point matters. Disney, by contrast, recently opted for a licensing arrangement with OpenAI, providing it with access to select character IP in exchange for an investment. Netflix is keeping this one internal.

The labor context cannot be ignored

The acquisition comes at a sensitive moment in Hollywood’s labor relations. Above-the-line unions are entering a new round of contract negotiations with studios and streamers, and AI is expected to be a central sticking point again. The 2023 strikes produced landmark language around consent and AI use, but enforcement and scope remain contested. Affleck is himself a signatory to the Creators Coalition on AI, which describes its mission as committed to “responsible, human-centered innovation” and explicitly states: “This is not a full rejection of AI. The technology is here.”

Netflix has not disclosed a timeline for when InterPositive’s tools will be available to its creative partners, nor has it provided specifics about which productions will have early access. The question worth watching is whether the artist-first framing that Affleck built into the company’s founding logic survives the process of being scaled inside one of the world’s most data-driven entertainment businesses.

What is your read on this deal? Let us know in the comments below!