Imagine watching a classic film where a long-gone star suddenly delivers brand-new dialogue in their own unmistakable voice, leaving you unsure whether the performance was ever truly theirs. That unsettling possibility now sits at the heart of Hollywood’s growing confrontation with artificial intelligence, a force reshaping how movies, shows and music get made while raising urgent questions about who owns a performer’s likeness and what counts as authentic creative work.
This article examines the ethical questions raised by AI tools in entertainment, from consent and job security to bias in representation and the shifting idea of authorship. It draws on real industry events, union agreements and specific productions to show how these technologies are already changing the landscape while highlighting the safeguards creators and regulators are trying to put in place.
The Rise of AI in Entertainment: A Double-Edged Sword
AI entered entertainment quietly at first through editing suites and concept-art generators before moving into more visible roles such as de-aging actors and creating background crowds. By 2024 studios like Disney and Warner Bros were testing these systems on major projects, and industry estimates suggested production costs could drop by as much as thirty per cent in certain departments. The savings matter because they influence decisions about which stories get told and who gets hired to tell them, yet the same tools also concentrate power in the hands of those who control the data and the algorithms.
Visual effects pipelines have seen some of the fastest change. What once required teams of artists for weeks can now be handled in days with machine-learning assistance for tasks like rotoscoping. Films such as Avatar: The Way of Water demonstrated how these methods could simulate complex natural phenomena, yet the underlying training data often comes from existing artist work without clear permission. Lawsuits against platforms like Midjourney illustrate the legal uncertainty that still surrounds ownership of the material feeding these models. Studios have quietly expanded these experiments into 2025 and 2026, with tools now assisting in crowd simulation for large-scale epics and even suggesting lighting setups that once demanded days of on-set tweaking.
Deepfakes: Consent and the Spectre of Exploitation
Deepfake technology sits at the most uncomfortable edge of this shift because it can replicate a person’s appearance and voice with unsettling accuracy. High-profile cases involving non-consensual imagery have already prompted public statements from actors such as Scarlett Johansson, and the issue extends into authorised uses as well. A fan-made trailer for The Lord of the Rings that recreated Ian Holm’s voice as Bilbo Baggins showed how fans might welcome such recreations, yet it also sparked discussion about whether estates or studios should control the digital legacy of performers who are no longer alive.
Regulators have begun to respond. The European Union’s AI Act treats deepfakes as high-risk applications that require clear labelling, while Hollywood unions push for stronger right-of-publicity laws. A 2024 Variety survey found that sixty-eight per cent of actors expressed concern about digital replicas replacing or diluting their work. Without consistent rules, the technology risks turning personal identity into something that can be licensed or sold without ongoing input from the original performer. Recent 2025 court rulings in the United States have started to test these boundaries, with several estates winning preliminary injunctions against unauthorised voice models trained on archival recordings.
Job Displacement: When Machines Write the Script
Writers and below-the-line artists feel the pressure most directly. Tools that analyse scripts for commercial potential or generate dialogue drafts have moved from experimental to practical use, and the 2023 Writers Guild strike placed explicit limits on AI involvement at the centre of negotiations. The fear is that studios could treat machine output as a first draft and then pay human writers only to polish it, lowering both compensation and creative ownership over time.
Voice actors face similar questions as synthetic systems replicate accents and emotional range with increasing fidelity. Entry-level positions in visual effects have already been affected by automation of repetitive tasks, contributing to the wave of layoffs that followed the pandemic. At the same time, some directors argue that AI can remove drudgery and free people to focus on the elements that require genuine human insight. Data from McKinsey suggests around twenty per cent of media jobs could be exposed to significant change by 2030, a figure that underscores why unions continue to press for transparency and residual payments tied to any digital replica. By early 2026, several major agencies began requiring clients to sign addendums that explicitly address how their digital likeness might be reused across future projects.
Bias and Representation: AI’s Hidden Prejudices
Because these systems learn from existing internet content, they tend to reproduce patterns already present in that data. Early facial-recognition tests showed higher error rates for people of colour, and music-generation models often favour Western musical structures. Research from the Geena Davis Institute in 2023 found that AI-created characters leaned heavily male and white, echoing long-standing imbalances in on-screen representation. These tendencies matter because they shape what audiences see and, over time, what stories studios consider commercially viable.
Netflix’s recommendation engine has drawn criticism for narrowing viewer exposure to independent or international titles, creating feedback loops that reinforce existing tastes. Calls for fairness audits and more diverse training sets have grown louder, yet meaningful enforcement remains inconsistent across companies. The result is a risk that cultural output becomes more uniform rather than more inclusive. Independent researchers released updated studies in 2025 showing that certain generative tools still under-represent non-English languages and regional storytelling traditions, prompting fresh calls for community-led data curation.
Case Studies: AI in Action and Backlash
Robert Zemeckis’s 2024 film Here used AI-assisted de-aging on Tom Hanks and Robin Wright to span multiple decades in a single narrative. The technical results drew praise for seamlessness, but some viewers noted an unsettling quality that echoed earlier experiments with digital rejuvenation. Producer Gary Goetzman described the process as a supporting tool rather than a replacement, yet the production still fed into broader conversations about how far such techniques should extend.
A separate flashpoint came from an AI-generated trailer for a hypothetical The Last of Us movie that spread rapidly online. While the marketing impact was clear, the project bypassed traditional union labour, prompting renewed discussion about whether AI-generated material should carry credits or trigger residual obligations. In music, the 2023 track Heart on My Sleeve that imitated Drake’s voice led to swift platform removals and legal action, exposing gaps in current intellectual-property protections. International examples, such as Bollywood studios using AI for dubbing regional films into Hindi, show both the accessibility gains and the questions of cultural authenticity that accompany wider adoption. Similar dubbing experiments reached Korean and Latin American markets by 2025, raising parallel debates about accent preservation and performer compensation.
Industry Responses and Regulatory Horizons
Studios have moved at different speeds. Universal Pictures signed an ethics agreement in 2024 that emphasises human oversight, while independent filmmakers often turn to open-source models for cost reasons. SAG-AFTRA’s AI Covenant now requires disclosure and compensation structures for any digital replica, giving performers clearer contractual footing than existed before the strike.
Legislation is evolving as well. California’s AB 1836 restricts unauthorised deepfakes in political contexts and has inspired similar proposals aimed at entertainment. China’s regulatory approach prioritises alignment with state values, creating a contrast with the more fragmented American framework. Voices such as researcher Timnit Gebru have argued for ethics boards that include technologists, artists and legal experts working together inside studios rather than after problems arise.
“AI is not the author; it’s the apprentice. The ethical path lies in symbiosis, not subjugation.” — Filmmaker Boots Riley, 2024 TED Talk
Future Outlook: Towards Ethical Innovation
Technical proposals such as blockchain provenance tracking and visible watermarking aim to make AI-generated content easier to identify. Advances in explainable AI could also help creators understand why a model produced a particular suggestion, reducing the sense that decisions arrive from a black box. Forecasts range from optimistic predictions of AI-assisted Oscar contenders by 2030 to warnings about an overwhelming volume of low-effort content crowding out original work.
Audience attitudes appear to be shifting as well. Recent polling indicates more than half of viewers would welcome clear labelling on projects that rely heavily on generative tools. At its best, the technology could help smaller teams create material in languages with limited resources or reconstruct historical settings for documentaries with greater accuracy. Energy consumption remains a practical concern, however, because training large models requires substantial power that conflicts with the sustainability messaging found in films such as Don’t Look Up. At Dyerbolical we have followed these developments closely because they touch every part of how stories reach audiences, and the central challenge is ensuring that efficiency gains do not come at the permanent expense of the human judgment that gives art its lasting value. You can read more about our approach at https://dyerbolical.com/about-us/.
Bibliography
SAG-AFTRA. “AI Protections in the 2023 TV/Theatrical Contract.” sagaftra.org, 2023.
McKinsey & Company. “The Economic Potential of Generative AI: The Next Productivity Frontier.” 2023.
Variety. “Actors Fear AI ‘Performance Doubles’ in Hollywood Survey.” 2024.
Geena Davis Institute on Gender in Media. “AI and Gender Bias in Visual Content.” 2023.
European Parliament. “The EU Artificial Intelligence Act.” Official Journal of the European Union, 2024.
California Legislative Information. “AB 1836: Deepfakes in Elections.” 2024.
TED. “Boots Riley on AI and Creative Labour.” 2024.
University of Southern California, Signal and Image Processing Institute. “Watermarking and Provenance Standards for Generative Media.” 2025 update.
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
