Inside John Lennon: The Last Interview – A 10% AI-Built Documentary Powered by Meta’s Video Models

The Meta Deal: Financing and Tooling in One Package

The structure of the partnership is worth understanding precisely, because it is not a simple licensing arrangement. Meta supplied both the generative-video tools used for the AI segments and the completion financing that allowed Soderbergh to finish the film. The deal runs parallel to Meta’s new multi-year Festival de Cannes sponsorship, which this year replaced TikTok as the festival’s headline tech partner.

This dual role — funder and toolmaker — is a configuration the industry has not seen at this scale before. It raises a question that no one at Cannes answered cleanly: where does production financing end and platform promotion begin?

Sean Lennon’s blessing letter was read aloud by festival director Iris Knobloch before the screening. Soderbergh, in a small but deliberate piece of theatre, chose to remain seated rather than stand to introduce the film himself.

What the AI Actually Does in the Cut

Soderbergh has been more precise than most filmmakers about where generative AI sits in his timeline. Speaking to Filmmaker magazine, he described the AI segments as “thematically surreal images that occupy a dream space rather than a literal space.” The material amounts to approximately ten minutes spread across the film’s 90-minute body, appearing primarily when Lennon and Ono drift into philosophical territory.

He framed the AI work as the final stage of finishing — prompted by “images that are impossible to shoot” combined with a project that had “run out of time and money” before Meta’s offer arrived.

That framing carries real weight for working professionals. Several outlets described the AI use as a quick substitute for traditional VFX. Soderbergh’s own words tell a different story: this was a budgetary and temporal impossibility, not a swap for a year’s worth of compositing work. The distinction matters enormously for any DP, editor, or VFX supervisor trying to map this tooling onto an actual production schedule.

The Ethical Framework Soderbergh Applied

In his Deadline Q&A, Soderbergh laid out a three-part test he applied to every AI shot:

• Is it necessary?
• Is it the only way to accomplish what I want to see?
• Is it truly the best way to do it?

He told the Associated Press he had no illusions about the reception: “I knew what was coming.” On the question of transparency, he went further:

“In the world outside of the creative context, we’re not aware of the extent that this is being used and used to manipulate us. I’m like my own whistle blower.”

That self-designation is significant. Soderbergh is not positioning himself as a pioneer celebrating a new frontier. He is positioning himself as someone who used a tool, disclosed it fully, and is now making the case that disclosure itself is the ethical act — not abstention.

A Festival Speaking in Two Voices

Cannes 2026 was always going to be a stress test for the festival’s posture on generative AI, and the Soderbergh screening sat awkwardly between two louder positions taken in the same week.

At the Théâtre Debussy, Guillermo del Toro introduced a 20th-anniversary Pan’s Labyrinth screening with a blunt “Fuck AI!” Peter Jackson, accepting his Honorary Palme d’Or, took the opposite position — telling the audience he would be open to AI for performance work, including digitally resurrecting performers, provided the guardians of their legacy signed off.

Inside the Marché du Film, the conversation was more procedural. At the AI for Talent Summit, Darren Aronofsky framed generative AI as a tool in the tradition of filmmaking evolution: “From when sound was first introduced, there was incredible pushback from all the people playing usable instruments. When the portable camera came, we suddenly got films like Breathless and the French New Wave.”

Three major filmmakers. Three incompatible positions. One festival week.

The Policy Gap Cannes Has Not Closed

Festival director Thierry Frémaux and Iris Knobloch articulated, at a Pathé Palace press conference earlier in the year, a curatorial exclusion of films “primarily driven by generative AI” from competition. Frémaux also floated the idea of a visible “made without artificial intelligence” label on selected works. Neither position has been codified as a written rule in the published 2026 regulations.

John Lennon: The Last Interview screened as a Special Screening rather than in any competitive section — neatly sidestepping the policy question entirely. Whether that placement was strategic or simply a function of the film’s subject matter, the effect was the same: the most prominent AI-assisted feature at Cannes 2026 existed in a regulatory grey zone the festival has not yet resolved.

The gap between del Toro’s expletive and Soderbergh’s Meta partnership is the gap Cannes will need to close — in writing, not in press conferences — before the next edition.

The Labor Question, From Someone Who Has Used the Tools

The Soderbergh comment most likely to resonate with working crews came in his conversation with The Next Web. He argued that “most jobs that matter when you’re making a movie cannot be performed by this tech and never will be performed by this tech.” He then added a craft-level observation that cuts directly against the standard generative-AI marketing pitch: “As it becomes possible for anybody to create something that meets a certain standard of technical perfection, then imperfection becomes more valuable and more interesting.”

That is not a reassurance designed to placate unions. It is a structural observation about how creative markets respond to commoditization. When technical execution becomes universally accessible, the premium shifts to judgment, taste, and intentional imperfection — qualities that remain stubbornly human.

Whether that dynamic plays out as Soderbergh predicts, or whether it simply compresses the market for mid-tier technical work, is the question the industry has not yet answered.

What This Means for AI Tool Adoption in Film Production

For founders and product teams building in the generative video space, the Soderbergh case establishes a reference point worth studying carefully.

The workflow that emerged here was not AI-first. It was AI-as-last-resort — applied at the finishing stage, under budget and time constraints, to material that had no viable alternative production path. That is a meaningfully different use case than the “AI replaces pre-production” narrative that dominates most product marketing.

The three-question framework Soderbergh applied — necessity, exclusivity, optimality — is also a practical evaluation rubric that any production team could adapt. It shifts the conversation from “can we use AI here?” to “should we, and can we defend that decision publicly?”

Meta’s dual role as toolmaker and financier is the structural detail that deserves the most scrutiny going forward. As AI video platforms seek distribution partnerships and content deals, the line between infrastructure provider and creative stakeholder will continue to blur. Productions that accept this kind of integrated deal will need to think carefully about what disclosure obligations come attached.

The Disclosure Standard Has Been Set

A release plan for John Lennon: The Last Interview has not been announced. The film remains in the hands of its producers and Meta for the next phase of its rollout.

What has been established, regardless of how the film performs commercially, is a disclosure standard. Soderbergh named the tool, named the percentage, named the funder, and named the ethical framework he applied — all before the lights went down. That level of specificity is rare in an industry where AI use is far more widespread than public acknowledgment suggests.

The more interesting question is not whether ten percent AI content is acceptable in a documentary. It is whether the industry will treat Soderbergh’s transparency as a model to follow — or as an anomaly that made everyone else’s silence more conspicuous.