Top AI Product

Netflix has never been big on acquisitions. In its 28-year history, the company has made only a handful of notable purchases. So when Bloomberg reported on March 11 that Netflix could pay up to $600 million for InterPositive — a 16-person AI filmmaking startup founded by Ben Affleck — the entertainment and tech worlds both took notice. If the full earn-out is realized, this would rank as Netflix’s second-largest acquisition ever, trailing only the $700 million Roald Dahl Story Company deal.

But this isn’t a typical AI hype story. InterPositive doesn’t generate videos from text prompts. It doesn’t scrape the internet for training data. And it explicitly cannot manipulate or replace human performances. So what exactly is Netflix paying half a billion dollars for?

What InterPositive Actually Does

InterPositive operates in a space most AI coverage ignores: post-production workflows.

Here’s how it works. The system takes a production’s existing dailies — the raw footage shot on set — and builds a custom AI model specific to that project. The model learns the visual language of the production: its lighting signatures, lens characteristics, and editorial logic. Filmmakers can then use this model during post-production to perform tasks that traditionally require expensive and time-consuming VFX pipelines.

Specific capabilities include:

• Relighting shots after they’ve been filmed
• Removing stunt wires and rigging
• Reframing to correct composition errors
• Recovering missed shots that would otherwise require reshoots
• Enhancing backgrounds without building full CG environments
• Color grading and mixing with AI-assisted precision

As Affleck put it: “It’s not about text-prompting or generating something from nothing. AI, people mostly think of it as making something from nothing. That’s not what this is.”

To build the foundational 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. This means InterPositive’s training data is entirely first-party — no scraped internet content, no unlicensed footage.

The Ben Affleck Factor

The story behind InterPositive is unusual. Affleck, best known as an actor and director (Good Will Hunting, Argo, The Town), quietly founded the company in 2022 and kept it in stealth mode for nearly four years. The 16-person team includes engineers, researchers, and creative executives who built the technology from scratch.

Affleck’s involvement isn’t just a celebrity endorsement play. As someone who has directed multiple feature films, he understood the specific pain points of post-production — the costs of reshoots, the limitations of traditional VFX, the frustration of losing usable footage to fixable technical problems. InterPositive was built by someone who actually sits in the editing room.

Under the deal terms, the entire InterPositive team joins Netflix, and Affleck takes on a role as senior adviser. Netflix plans to offer InterPositive’s technology to its creative partners — the directors, showrunners, and producers working on Netflix originals — but has no plans to sell it commercially. This is a proprietary competitive advantage play, not a SaaS product.

How InterPositive Compares to Other AI Film Tools

The AI filmmaking space is crowded, but most players are solving different problems than InterPositive.

Runway is the most well-known name in AI video. Its Gen-4 models focus on generative capabilities — creating video from text prompts, image-to-video conversion, and general creative tools. Runway targets a broad audience from social media creators to professional studios. It raised $141 million at a $4 billion valuation. But Runway’s core product is generative: it creates new content from scratch.

Wonder Studio (by Wonder Dynamics, acquired by Autodesk) specializes in CG character replacement. It takes live-action footage, rotoscopes actors out, and replaces them with 3D animated characters — useful for VFX-heavy productions but narrow in scope.

Sora (OpenAI) and Veo (Google DeepMind) are text-to-video models designed for content generation. They’re impressive technical achievements but address a fundamentally different use case than what InterPositive does.

InterPositive’s differentiation is clear: it doesn’t generate new content. It enhances existing footage using models trained exclusively on that production’s own material. This makes it more of a post-production workflow tool than a creative AI, which is precisely why it appeals to Netflix. The streaming giant needs to produce hundreds of titles per year efficiently, and shaving time and cost off post-production at scale is worth far more than any text-to-video generator.

The $600 Million Question: Is It Worth It?

The headline number — $600 million — needs context. According to Bloomberg’s reporting, the upfront cash payment was significantly less than $600 million. The full amount is an earn-out, contingent on InterPositive meeting performance targets after integrating into Netflix. This structure is common in tech acquisitions and reduces Netflix’s risk.

Still, the valuation is aggressive for a 16-person startup. That works out to roughly $37.5 million per employee at the maximum earn-out. For comparison, Google’s acquisition of DeepMind in 2014 valued each of the roughly 75 employees at about $6.7 million.

Netflix’s logic likely comes down to scale economics. The company spent an estimated $17 billion on content in 2025. If InterPositive’s technology can reduce post-production costs by even a few percentage points across Netflix’s slate of hundreds of productions per year, the acquisition pays for itself quickly.

There’s also a defensive angle. By acquiring InterPositive outright and keeping the technology proprietary, Netflix ensures competitors like Disney+, Amazon Prime Video, and Apple TV+ can’t access the same tools. In a streaming war defined by razor-thin margins and relentless content volume, any production efficiency advantage compounds over time.

Hollywood’s Uneasy Relationship with AI

The timing of this acquisition is politically charged. Hollywood’s 2023 strikes — led by SAG-AFTRA and the WGA — were driven in large part by fears about AI replacing human workers in film and television production. The resulting contracts included specific provisions around AI use, and the next round of union negotiations is approaching.

InterPositive has clearly been designed with these tensions in mind. The system explicitly cannot be used to manipulate, replace, or synthesize human performances. It focuses on technical post-production tasks (relighting, wire removal, color grading) rather than anything touching actor likenesses or writer contributions.

Netflix’s chief product and technology officer framed the acquisition carefully, stating that InterPositive’s team shares Netflix’s belief that “innovation should empower storytellers, not replace them.”

Despite this messaging, the International Alliance of Theatrical Stage Employees (IATSE) — the union representing Hollywood’s technical workers, including many post-production professionals — declined to comment on the deal. The silence speaks volumes. Post-production workers, including colorists, VFX artists, and editors, are the ones most directly affected by tools like InterPositive. Whether this technology augments their work or gradually displaces it remains an open question.

FAQ

How much did Netflix pay for InterPositive?
Netflix could pay up to $600 million total, but the upfront cash amount was lower. The full $600 million is contingent on InterPositive meeting performance targets after integration — a standard earn-out structure in tech acquisitions.

Will InterPositive be available to filmmakers outside Netflix?
No. Netflix has stated it will offer InterPositive’s technology to its own creative partners — directors, producers, and showrunners working on Netflix content — but has no plans to sell or license it commercially.

Does InterPositive use generative AI to create new video content?
No. Unlike tools such as Runway or Sora, InterPositive does not generate content from text prompts or create new footage. It builds AI models from a production’s existing footage and uses those models to enhance shots during post-production — relighting, wire removal, reframing, color grading, and similar tasks.

Can InterPositive alter actor performances or likenesses?
No. The system is specifically designed so it cannot be used to manipulate, replace, or synthesize human performances. It focuses exclusively on technical aspects of post-production.

Who founded InterPositive and when?
Ben Affleck founded InterPositive in 2022. The company operated in stealth mode until Netflix announced the acquisition on March 5, 2026. Affleck will serve as a senior adviser to Netflix as part of the deal.

You Might Also Like

• Llm Skirmish What Happens When you let ai Models Fight Each Other in an rts Game
• Sakana ai doc to Lora Text to Lora Your llm Just got a Permanent Memory Upgrade
• Saguaro Speculative Speculative Decoding the yo Dawg i Heard you Like Speculation Approach to Faster llm Inference
• Tropes fyi llm Writing Tropes a Brutally Honest Catalog of how ai Gives Itself Away
• Your Anonymous Posts Arent Anonymous Anymore Inside the llm 大规模去匿名化研究