Higgsfield's 'The Cully Hill Boys': The First Fully AI-Generated Feature Film Goes Mainstream

For roughly three years, “AI video” meant eight-second clips of distorted hands and uncanny lip-sync — party tricks, not premieres. That era ended on August 5, 2026, at The Glasshouse in New York City, when AI video startup Higgsfield AI premiered The Cully Hill Boys: a 110-minute action-comedy in which every frame was generated by artificial intelligence. Now the film is drawing fresh national attention — Semafor ran a full review on August 19 — and the conversation has shifted from “can AI make a movie?” to “what happens to the movie business now that it can?”

What premiered in New York

The Cully Hill Boys follows three struggling East London rappers whose plan to shoot a music video unravels when a stolen boat turns out to be stuffed with cash, landing them between rival crime cartels. The premise is unremarkable. The production is not: the entire film — every shot, every location, every performance — was generated with AI, then assembled and finished by a human post-production team.

The digital cast consists of four real entertainers who licensed their likenesses to the production: two-time UFC champion Israel Adesanya, MMA veteran and actor Quinton “Rampage” Jackson, livestream creator N3on, and streetball athlete Matt Kiatipis. Each performer provided high-resolution photographs and voice recordings; AI models then rendered them “acting” throughout the film’s hundred-plus minutes.

Semafor’s critic came away impressed but clear-eyed. The film has moved far beyond the infamous 2023 clip of Will Smith eating spaghetti — characters have the correct number of fingers — but it retains “at times, an unrealistic over-polished sheen.” The strongest suit, notably, was the writing, which is not an AI accident: Higgsfield tapped a human screenwriter, Timothy W. Planagan, who spent roughly five years on the screenplay. It became a Feature Screenplay Finalist at the 2026 London Independent Screenplay & Film Awards and appeared on The Black List — Hollywood’s annual survey of most-liked unproduced screenplays — where Higgsfield discovered it and licensed the rights for an AI adaptation.

The numbers that made studios pay attention

Higgsfield shared production economics with Forbes after the premiere, and they are the real story:

• Total budget: approximately $2 million, of which about $1 million went to AI compute (tokens and generation infrastructure)
• Production time: four weeks from start to finish
• Scale: roughly 1,000 production-ready assets across more than 100 locations, all in 4K
• Team: a 28-person crew, of which around 40% had a traditional film background and no prior AI experience, onboarded through Higgsfield’s free Academy curriculum
• Process: nine directors developed sequences concurrently under emerging filmmakers Aitore Zholdaskali and Adilet Abish

For comparison, Higgsfield’s previous feature, Hell Grind — a 95-minute AI film launched at Cannes — cost about $500,000 and required roughly 100 assets. By that measure, the company multiplied its production scale roughly tenfold in a matter of months. Independent films of comparable ambition, Higgsfield notes, typically demand one to two years and budgets above $20 million. Those comparison figures come from the company itself, and deserve skepticism — but even with generous discounting, the cost structure is one no studio CFO can ignore.

Notably, the film is not headed for a conventional theatrical run. Higgsfield built it primarily to demonstrate its flagship product, Cinema Studio 4.0, to studios, agencies, and brands. The movie is, in effect, a very expensive — by startup-demo standards, shockingly cheap — sales artifact.

Consent as commercial architecture

The most consequential innovation of the premiere never appeared on screen. It was written into the contracts.

Each cast member’s agreement covered compensation, permitted scope of use, and script approval before production began — every performer reviewed the full screenplay and their character profile. Higgsfield says all source assets (the photos and voice recordings) were permanently deleted within 30 days of wrap, and that its agreements bar the company from retaining, training on, or licensing that data to third parties.

This operationalizes principles the industry has spent three years fighting to establish. The 118-day SAG-AFTRA strike of 2023 produced Hollywood’s first AI consent provisions, requiring advance notice and compensation before a performer’s digital replica can be used. Congress is now moving to give that principle the force of law via the NO FAKES Act of 2026 (S. 4591), which would create a federal intellectual-property right over every individual’s voice and visual likeness, enforceable against unauthorized replicas.

What labor won at the bargaining table and lawmakers are writing into statute, The Cully Hill Boys converts into commercial practice. Since AI imagery became photorealistic, a performer’s likeness has been something to defend; in this film, it became something to license. The market is already organizing around the shift: according to the premiere announcement, the agencies representing N3on and Kiatipis now offer AI likeness licensing alongside sponsorships and endorsements — treating a client’s AI avatar as a managed line of business rather than a standing vulnerability.

What remains human

Post-production — led and executed entirely by humans — absorbed a substantial share of the four-week schedule. AI generation delivered raw material; dedicated photo and film editors refined generated frames to preserve continuity across shots, sequences, and locations. The directing team held final editorial authority, and with nine directors working simultaneously, that human editorial layer is what gave the film a single visual and narrative style. Choices about voices, accents, sound effects, music, and wardrobe remained, as in traditional production, human decisions.

Semafor’s review captures the residual gap in one sharp line: in a scene where a character is violently beaten, “his reaction is overly nonchalant” — a moment of apparent prompt fatigue. “In the case of AI films,” the critic writes, “poor prompting is just the equivalent of bad acting.”

Why it matters

Three takeaways for anyone tracking generative AI’s march through the creative industries:

1. The capability curve is steepening, not flattening. Tenfold production-scale gains in months mean the “AI feature film” is leaving the novelty zone. Studios are no longer asking for proof of concept; they’re asking for unit economics.
2. Consent frameworks are becoming product features. The industry’s scarcest resource is no longer compute — it’s trust. Higgsfield’s deletion guarantees, script approval, and scope limits are as much a go-to-market strategy as an ethical stance, and competitors will have to match them.
3. The bottleneck is moving up the stack. With image generation commoditizing, the differentiators are story, editorial judgment, and rights management — precisely the layers that stayed human in this production. Screenwriters with unproduced Black List scripts may have just found an unexpected buyer.

Higgsfield CEO Alex Mashrabov frames the company’s bet plainly: “By removing physical constraints and reducing production costs, we’re empowering filmmakers, creators, and studios to unlock a whole new talent pool and tell stories that were previously out of reach.”

Whether The Cully Hill Boys endures as a film is a verdict audiences will render in time. The precedent it establishes — every face duly authorized before being delivered through a brand-new storytelling technology — will outlast that verdict. Industries rarely retreat from cost structures and consent frameworks that have proven to work.