Picture an actor who passed away years ago delivering a fresh line reading in a major studio film, every inflection drawn from old recordings and stitched together by code. That scenario sits at the center of this article, which examines the growing use of artificial intelligence to recreate performers in Hollywood productions, the labor conflicts it has triggered, and the deeper questions it raises about identity, consent, and what counts as authentic performance.
Technology has long altered how stories reach the screen, yet the current wave of generative tools moves beyond simple visual effects into the territory of full performance synthesis. Studios can now train models on existing footage and audio to produce new scenes featuring actors who never consented to the work or who are no longer alive to object. The result is a collision between commercial efficiency and long-standing protections for performers, one that echoes earlier cultural anxieties about capturing or mimicking the dead.
Historical Context: From VFX to Digital Ghosts
Hollywood began experimenting with digital stand-ins decades ago. In the 1990s, films like Forrest Gump used early compositing methods to place John F. Kennedy in new scenes, relying on archival material and careful editing. Those techniques required extensive human labor and still looked imperfect up close. Modern generative systems such as Deep Voodoo and Runway ML shift the process further by creating entire performances from limited source material, reducing the need for additional shooting or stand-ins. The 2019 short film He’s Expecting demonstrated this capability by bringing Paul Walker back for a brief appearance after his death during production of Fast & Furious 7. While some viewers appreciated the sense of closure, the project also highlighted the absence of any prior agreement from the actor about how his likeness might be reused.
These capabilities invite comparison with earlier attempts to bridge the gap between the living and the departed. Spirit photography in the late 1800s promised images of loved ones who had passed, yet many examples turned out to rely on double exposures and other darkroom tricks. The technology fooled audiences because people wanted the evidence to be real. AI-generated performances create a similar tension today: the images and voices can appear convincing enough to suspend disbelief, yet the underlying process remains a simulation rather than a continuation of the original person.
Key Milestones in AI Film Integration
The shift from enhancement to outright replacement appears in several documented cases. In 2016, Rogue One employed motion capture to de-age a younger version of Carrie Fisher, testing the limits of what could be altered without replacing the performer entirely. Five years later, a fan project titled Dragon Slayer placed a CGI version of James Dean into a Back to the Future scenario, prompting immediate pushback from representatives of his estate who argued the use violated the actor’s legacy. By 2023, the SAG-AFTRA strike elevated these concerns to contract negotiations, with performers demanding explicit rights over any digital replica of their work.
Each step shows how quickly the tools moved from correction and de-aging toward generating new material without the original artist’s involvement. That trajectory matters because it changes the economic relationship between performers and studios, turning an actor’s past output into reusable training data rather than a one-time performance.
The 2023 Strike: Epicentre of the Controversy
The 2023 SAG-AFTRA strike lasted 118 days and centered in part on protections against unauthorized digital replicas. More than 160,000 performers participated, seeking guarantees that their likenesses could not be scanned and reused indefinitely without fresh consent or compensation. Studio proposals at the time suggested that training data could be collected after an actor retired, with no requirement for ongoing payments. Duncan Crabtree-Ireland, the union’s chief operating officer, described the risk as a potential death of performance, where algorithms could generate new scenes from existing material without ever hiring another human actor.
Actress Fran Drescher, serving as union president, compared the unchecked spread of these tools to Frankenstein’s monster, a creation that escapes its makers’ control. On the picket lines, members shared stories of studios already testing AI-generated crowds for background scenes in an effort to cut costs. One visual effects artist working on a horror project recalled producing a digital horde of figures that looked indistinguishable from real extras, noting how the result raised questions about whether any human presence remained necessary. The final agreement included requirements for consent and compensation when replicas are used, yet questions persist about how those rules will be monitored as the technology continues to advance.
Ethical and Philosophical Underpinnings
Contract language only addresses part of the unease. Recreating performers such as Marilyn Monroe or Robin Williams without family approval raises concerns about the permanent alteration of an artist’s public record. Philosopher Nick Bostrom’s simulation hypothesis offers one framework for thinking about these recreations: if digital versions can behave convincingly enough, viewers may begin to treat them as equivalent to the original person. In the broader cultural conversation, this possibility intersects with older ideas of thought forms or tulpas, entities brought into being through sustained collective attention. Film scholar Jason Mittell has pointed out that such synthetic performances risk flattening the unpredictable human elements that give stories their emotional weight.
Legal Battles and Precedents
California’s AB 1836, passed in 2024, prohibits unauthorized digital replicas of deceased performers within the state, though federal gaps remain that could allow similar uses elsewhere. Scarlett Johansson filed suit after an AI chatbot replicated her voice without permission, illustrating how the same tools extend beyond film into interactive media. Updated union contracts now require watermarking of AI-generated content so audiences can distinguish it from traditional performances. These measures aim to preserve a record of what was created by human actors versus what was synthesized later, much as investigators once sought ways to verify whether spirit photographs contained genuine anomalies or simple manipulation.
AI in Paranormal and Horror Genres: Heightened Stakes
Genres that trade in atmosphere and the unexplained face particular complications. The Conjuring series, for example, depends on performances that feel grounded even when depicting extraordinary events. Directors such as James Wan have acknowledged the efficiency gains from AI tools while expressing concern that the results can lack the lived-in quality audiences respond to. Robert Zemeckis’s 2024 film Here used AI de-aging on Tom Hanks and Robin Wright to span multiple decades of a single story. Reviewers noted the technical achievement but also remarked on an emotional distance in the performances that traditional aging makeup or casting choices might have avoided.
In projects involving cryptids or historical mysteries, the same tools could generate footage of events that never occurred with enough detail to complicate ongoing debates. Some observers wonder whether training data drawn from real investigations might accidentally preserve fleeting anomalies, while others worry that fabricated evidence could further muddy already contested records. Horror scripts drafted partly by AI have begun to feature malevolent algorithms as central threats, creating a self-referential loop where the industry’s own tools become the subject of its stories.
Theories and Future Implications
Observers have outlined several overlapping reasons for the resistance. Economic displacement tops the list, as background performers and even supporting roles become easier to replace. Existential concerns follow, since digital recreations challenge traditional views of mortality and legacy in ways reminiscent of 19th-century spiritualist movements. A third worry centers on creative narrowing: models trained predominantly on existing works may favor repetition over genuine innovation. Emerging responses include blockchain-based contracts that could verify human contributions and broader ethical guidelines for AI use in entertainment. The underlying uncertainty remains whether these systems will ultimately expand storytelling options or simply reproduce familiar patterns at lower cost.
At Dyerbolical we have tracked how similar questions of authenticity appear across different fields of investigation, from historical footage analysis to contemporary claims of unexplained phenomena. The Hollywood debate supplies a clear case study in how quickly new tools can outpace the social agreements meant to govern them.
Conclusion
The controversy over AI in Hollywood extends beyond a single labor dispute. It forces a sustained look at how performances are valued, who controls an artist’s image after the fact, and what audiences accept as real within fictional worlds. Digital resurrections bring earlier cultural fascinations with the boundary between life and death into a commercial setting, where the stakes involve both creative integrity and basic compensation. As the technology matures, the outcomes will depend on whether consent mechanisms keep pace with what the tools can produce. The questions are unlikely to resolve quickly, yet they continue to shape decisions on sets and in contract rooms alike.
Bibliography
SAG-AFTRA, “2023 Strike Agreement Summary,” 2023.
California Legislative Information, “AB 1836 Digital Replica Protections,” 2024.
Nick Bostrom, “Are You Living in a Computer Simulation?,” Philosophical Quarterly, 2003.
Jason Mittell, Complex TV: The Poetics of Contemporary Television Storytelling, 2015.
Robert Zemeckis, Here production notes and interviews, 2024.
James Wan, public statements on visual effects tools, 2023-2024.
William Mumler historical records on spirit photography practices, 1860s-1870s.
Scarlett Johansson v. AI chatbot developers court filings, 2024.
Got thoughts? Drop them below!
For more articles visit us at https://dyerbolical.com.
Join the discussion on X at
https://x.com/dyerbolicaldb
https://x.com/retromoviesdb
https://x.com/ashyslasheedb
Follow all our pages via our X list at
https://x.com/i/lists/1645435624403468289
