When Darren Aronofsky, the visionary filmmaker behind Requiem for a Dream, Black Swan, and The Whale, announced his latest project would employ artificial intelligence to generate historical imagery for a docudrama, the reaction from Hollywood and the broader creative community was swift and polarized. The project, which uses AI-generated visuals to reconstruct historical events for which no film footage exists, represents one of the most high-profile experiments yet in blending generative AI with traditional filmmaking. It also raises profound questions about authenticity, labor, and the future of visual storytelling in an era when the tools of creation are being fundamentally rewritten.
The project, as detailed by Ars Technica, centers on using AI image and video generation tools to depict scenes from history that were never captured on camera. Aronofsky has framed the endeavor not as a replacement for human artistry but as an extension of the documentary tradition — one that has always relied on reenactments, illustrations, and creative interpretation to fill gaps in the visual record. The director has argued that AI-generated imagery, when clearly contextualized, is simply the next evolution of these long-standing techniques.
A Filmmaker’s Rationale: Why AI, and Why Now?
Aronofsky’s reasoning, as he has explained in interviews, is rooted in both pragmatism and artistic ambition. Traditional historical documentaries have long faced a fundamental constraint: for events that predate the invention of photography or film, creators must rely on static paintings, written accounts, or staged reenactments with actors. Each of these methods carries its own set of compromises. Reenactments can feel stilted and artificial; paintings reflect the biases of their era; and written narration, however eloquent, lacks the visceral immediacy of moving images. Aronofsky has suggested that generative AI offers a new path — one that can produce photorealistic depictions of historical moments with a fluidity and detail that traditional methods cannot match.
According to the reporting by Ars Technica, Aronofsky has been careful to distinguish his approach from the kind of AI-generated content that has drawn widespread criticism — deepfakes, synthetic celebrity likenesses, and the wholesale replacement of human artists. His team has emphasized that the AI-generated visuals are being used to depict events and figures from the distant past, not to fabricate scenes involving living individuals. The project also reportedly involves significant human oversight, with artists and historians working alongside the AI tools to guide the output and ensure historical accuracy. Aronofsky has described the process as collaborative, comparing it to a director working with a visual effects team rather than a filmmaker ceding creative control to a machine.
The Backlash: Artists, Unions, and the Ethics of Generative Imagery
Despite Aronofsky’s framing, the project has ignited fierce debate. Critics within the film industry and the broader arts community have argued that using AI-generated imagery — regardless of the context — legitimizes a technology that threatens the livelihoods of illustrators, concept artists, VFX professionals, and actors. The Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) and the International Alliance of Theatrical Stage Employees (IATSE) have both been vocal in recent years about the risks AI poses to creative workers, and Aronofsky’s project has become a lightning rod for those concerns. For many, the issue is not whether AI can produce compelling images but whether it should be used when human artists could do the work instead.
The timing of the project is particularly fraught. Hollywood is still processing the aftershocks of the 2023 strikes by SAG-AFTRA and the Writers Guild of America, both of which featured AI protections as central bargaining issues. The contracts that ended those strikes included provisions governing the use of AI in film and television production, but many in the industry view those protections as incomplete and vulnerable to erosion. Aronofsky’s decision to employ generative AI in a high-profile production has been interpreted by some as a test case — a way for the industry to gauge how audiences and workers will respond to AI-assisted filmmaking at the prestige level.
Historical Precedent: Documentaries Have Always Fabricated
Supporters of the project, however, point to a long and often uncomfortable history of fabrication in documentary filmmaking. Robert Flaherty’s 1922 classic Nanook of the North, widely considered one of the first feature-length documentaries, staged many of its scenes and presented them as authentic. Errol Morris, one of the most celebrated documentarians of the modern era, has built much of his career on stylized reenactments. Ken Burns, whose work has defined the genre for American audiences, relies heavily on the “Ken Burns effect” — slow pans and zooms across still photographs — precisely because moving footage of the Civil War or the early 20th century does not exist. In this context, Aronofsky’s use of AI can be seen as a continuation of a tradition rather than a rupture.
What distinguishes the current moment, of course, is the scale and sophistication of the technology involved. AI-generated video has improved dramatically in recent years, with tools from companies like OpenAI, Runway, and others capable of producing footage that is increasingly difficult to distinguish from reality. This capability introduces new risks around misinformation and audience trust. If viewers cannot tell whether they are watching real footage, a reenactment, or an AI-generated simulation, the epistemological foundations of documentary filmmaking are called into question. Aronofsky and his team have reportedly committed to clear labeling of AI-generated content within the film, but critics argue that such disclosures are easily overlooked or forgotten by audiences.
The Technology Under the Hood
The specific AI tools being used in Aronofsky’s production have not been fully disclosed, but the Ars Technica report indicates that the project is leveraging state-of-the-art generative video models. These models, trained on vast datasets of images and video, can produce remarkably detailed and coherent visual sequences based on text or image prompts. The technology has advanced rapidly; just two years ago, AI-generated video was marked by obvious artifacts — distorted hands, flickering backgrounds, uncanny facial expressions. Today’s models, while still imperfect, are capable of producing output that can pass casual inspection, particularly when integrated into a professionally edited production with music, narration, and sound design.
The involvement of human artists in the pipeline is a key element of Aronofsky’s defense of the project. According to the reporting, historians and visual artists are deeply embedded in the production process, providing reference materials, reviewing AI outputs for accuracy, and making manual adjustments where necessary. This hybrid workflow — part human, part machine — mirrors the approach being adopted in other creative industries, from architecture to advertising. Proponents argue that it represents the most responsible way to integrate AI into creative work: not as a replacement for human judgment but as a tool that amplifies it.
Audience Trust and the Documentary Contract
At the heart of the controversy is a question about the implicit contract between documentary filmmakers and their audiences. When viewers watch a documentary, they bring a set of expectations about truthfulness and authenticity that differ from those they bring to fiction. Even when documentaries employ reenactments or stylized visuals, audiences generally trust that the underlying facts are being represented honestly. AI-generated imagery complicates this contract in new ways. If a viewer sees a photorealistic depiction of, say, a Roman senate debate or a medieval battle, they may unconsciously assign it the same evidentiary weight as actual footage — even if a disclaimer has been provided.
This concern is not merely theoretical. Research in cognitive psychology has consistently shown that visual information is processed differently from textual or auditory information; people are more likely to remember and believe claims that are accompanied by images, regardless of whether those images are authentic. In an era already plagued by misinformation and declining trust in media institutions, the introduction of AI-generated historical footage into the documentary form carries risks that extend well beyond the boundaries of any single film.
What Aronofsky’s Bet Means for the Industry
Regardless of how Aronofsky’s project is ultimately received, it is likely to serve as a bellwether for the broader integration of generative AI into filmmaking. If the film is critically acclaimed and commercially successful, it will embolden other filmmakers and studios to adopt similar techniques. If it is rejected by audiences or condemned by industry peers, it may slow — though almost certainly not stop — the adoption of AI in prestige productions. The technology is advancing too quickly and the economic incentives are too powerful for the genie to be put back in the bottle.
For now, the project stands as one of the most consequential experiments in the ongoing negotiation between human creativity and machine capability. Aronofsky, a filmmaker who has never shied away from provocation, appears to relish the debate. Whether his gamble pays off — artistically, commercially, and ethically — will depend not just on the quality of the finished product but on the willingness of audiences, critics, and the industry to grapple honestly with the questions it raises. As the tools of visual creation become ever more powerful and accessible, the answers to those questions will shape the future of storytelling itself.
WebProNews is an iEntry Publication