An essay circulating this month frames the AI film studio as a 2030 vision — a fortress about to fall. The actual 2026 record says the fortress is already being renovated from the inside, by the people who own it, with real capital and a real legal framework nine days old. This letter is about what's already priced, not what's still speculative.
A widely-shared essay this month argues that AI will do to Hollywood what esports did to traditional sports — build a parallel star system from nothing, discover a viral unknown in Lagos or Seoul, and render the old gatekeepers irrelevant, sometime around 2030. It's a compelling frame. It is also, as of this month, already overtaken by events. Google DeepMind did not wait for an unknown creator to go viral. On June 22, it took its first-ever equity stake in a film studio — $75 million into A24, the studio behind Everything Everywhere All at Once — to co-develop AI filmmaking tools from inside an active, Oscar-winning production house. The disruption the essay predicts from outside the fortress is instead being built by the people who already own it.
The DeepMind-A24 arrangement is deliberately narrow in a way that matters for how this plays out. It is non-exclusive, does not touch A24's existing film library, and is not structured as an IP-licensing or model-training agreement — the two organizations work side by side on active productions, with DeepMind researchers embedded to observe how real filmmakers actually use generative tools. That is the same pattern DeepMind has already run with director Darren Aronofsky's production company on short films including Ancestra, now applied at studio scale for the first time. The stakes are correspondingly higher: a research partnership with one arthouse studio's worth of active productions is a materially different data asset than a handful of festival shorts.
A24 is not the only major studio treating AI as core infrastructure rather than a side experiment. Lionsgate named its first chief AI officer this year and told investors it has deployed AI tools across more than 80% of its workforce. Amazon MGM built an AWS-based production platform called Project Nara that tracks the provenance of every AI-generated asset — a direct response to the industry's central fear, that generated content can't be traced back to what trained it. Netflix acquired Ben Affleck's AI filmmaking startup outright. None of these are AI-native upstarts storming a fortress from outside. They are the fortress, retooling itself.
The essay's star-discovery scenario — an AI performer built entirely from an uploaded face and voice, cast across dozens of productions simultaneously — now has to clear an actual legal bar that didn't exist a month ago. SAG-AFTRA's new TV/Theatrical Agreement, ratified June 4 with 91.42% support and effective July 1, permits AI-generated performers only when they provide what the contract calls "significant additional value" over a live actor or that actor's own digital avatar. That standard is untested and will likely be litigated project by project, but it is a real constraint the essay's frictionless star-system scenario doesn't account for. The 2026 WGA deal reached in April runs the same logic on the writing side: if a studio trains a model on a member's scripts, the writer must be compensated for it.
Disney's own arrangement with OpenAI shows the more contested version of this same boundary. Sora-generated fan shorts now draw on more than two hundred Disney, Marvel, and Pixar characters — a studio licensing its own IP into a generative system, rather than an outside creator building a new star from nothing. That is a fundamentally different economic structure from the essay's Lagos-uploads-their-face scenario: value accrues to whoever controls the underlying IP and likeness rights, not to whichever unknown performer the algorithm happens to surface.
The tool makers are courting creators directly. Runway funds AI-assisted films. Adobe stood up its own fund. Google is training the next generation of AI filmmakers through Sundance. Every major studio is choosing to build the parallel universe rather than let someone else build it around them.
PitchBook's own data on AI video and media startups shows venture investment already running at $5.6 billion for 2026, up more than 43% from all of 2025 — and that figure was recorded before the Google-A24 deal closed. Early money concentrated almost entirely in the underlying video-generation models themselves, Runway and Luma AI chief among them. The newer, smaller rounds — Californian studio Promise backed by Andreessen Horowitz and Google's own AI Futures Fund, London-based Wonder's $12 million raise from Atomico and LocalGlobe — are the first wave of capital explicitly betting on the studio layer described in the viral essay: AI-native production houses building an IP catalogue rather than licensing tools to others. Wonder's own framing to investors is explicit about where the value actually sits: the model layer is commoditizing, and the durable economics move to whoever owns the finished IP.
South Korea offers the clearest government-backed version of this same bet. Seoul tripled its national AI budget this year and earmarked roughly $5.4 million in emergency film-industry funding specifically for AI-assisted productions, on top of an 81% increase to the film sector's annual budget. Two fully AI-generated features already reached South Korean cinemas in May — a sci-fi courtroom drama and a period piece, the latter already drawing attention on the AI festival circuit. This is a live production pipeline, not a pilot program.
The essay's core prediction — that AI transforms who gets to make films and who gets to be a star — is directionally right and probably still understates the eventual scale. Where it's wrong is on mechanism and timing. It frames this as an outside-in disruption arriving around 2030, the way esports built its own star system independent of the NBA and NFL. What's actually happening in 2026 is inside-out: Google buying into A24 rather than funding a rival to it, Lionsgate and Amazon MGM building AI into core production infrastructure rather than losing ground to someone who does, and SAG-AFTRA writing the legal floor before a single AI-native star has actually broken out. The unknown-creator-goes-viral scenario remains genuinely possible — South Korea's government-funded pipeline is arguably the most likely place for it to happen first — but the more probable near-term outcome is that the major studios absorb AI production capability faster than any outside challenger can out-produce them on volume alone. The risk to this thesis: if the "significant additional value" standard in the new SAG-AFTRA contract gets interpreted narrowly by courts or arbitrators, AI-native performers may be constrained inside the studio system for years before any true outside disruption becomes commercially viable — a slower, more contested path than either the essay's timeline or this letter's own base case assumes.
Founder, NextGen Economics · Bangalore, India · July 2026
Sources: Forbes, "Hollywood Studios Are Spending On AI To Control The Future Of Film" (Jun 2026) · TechTimes / Republic World, Google DeepMind–A24 investment coverage (Jun 2026) · Yahoo Finance / PitchBook News, "From models to studios: how AI video investment is evolving" (Jun 2026) · CNN, "South Korea's movie industry is embracing AI" (May 2026) · AIinvasion.com, "AI in Movies 2025–2026" (Apr 2026). This letter responds to a widely-circulated independent essay on AI film studios; that essay is not an NGE publication and its projections are not independently verified here.
A deeper research note on AI-native studio valuations and the SAG-AFTRA "significant additional value" standard is available to NGE clients beyond what's covered in this free letter.
Not investment advice. This letter evaluates a media-industry structural trend; it does not constitute a recommendation regarding any security.