From Emergence to Ethical Frontiers: AI and Deepfake Technology's Dual Impact on Cinema (2017–2024)

Bulgarova Bella Akhmedovna ORCID: 0000-0001-6005-2505 PhD in Philology Associate Professor; Department of Mass Communications; RUDN University Miklukho-Maklaya str., 6, Moscow, 117198, Russia bulgarova-ba@rudn.ru Other publications by this author     Ullah Md Tauhid ORCID: 0009-0008-5740-0263 Graduate student; Faculty of Philology; RUDN University Miklukho-Maklaya str., 6, Moscow, 117198, Russia 1032229412@pfur.ru Mondal Papiya ORCID: 0009-0006-0337-7059 Postgraduate student; Faculty of Philology; RUDN University Miklukho-Maklaya str., 6, Moscow, 117198, Russia 1042235023@pfur.ru Other publications by this author

Ma Fei ORCID: 0009-0002-5688-1348 Postgraduate student; Faculty of Philology; RUDN University Miklukho-Maklaya str., 6, Moscow, 117198, Russia 1042238104@pfur.ru Other publications by this author

Introduction

Deepfake uses sophisticated computer algorithms known as generative adversarial networks (GAN) to create amazingly realistic media. Deepfake technology originated on the basis of Generative Adversarial Networks (GAN), developed by Ian Goodfellow in 2014. GAN has two neural networks - one is a generator and the other is a discriminator - that work closely together to generate outputs that become more realistic with each iteration. Synthetic data is generated by the generator, and the validity of this data is evaluated by the discriminator. This continuous process leads to the creation of very plausible fakes. When it comes to further development, AI mainly relies on machine learning (ML) and neural networks (NN) ^{[1, pp. 316-323]}. Based on extensive facial data sets, this technology is capable of creating ultra-realistic images that find universal application in the film industry, for example, to slow down the aging of actors, digitally resurrect deceased celebrities and create completely new characters. Technologies based on artificial intelligence have become able to accurately edit and reproduce human appearance and voice characteristics in audiovisual content, which leads to a gradual blurring of the boundaries between real and fake ^{[2, pp. 216-224]}. This evolution reflects historical changes in film technology: from silent films to sound and color, from analog to digital - each transformation has significantly changed both the art of storytelling and the perception of the audience. The development of deepfake technology, similar to such major innovations, will significantly revolutionize creative practice, the production process, and the realism of cinematic storytelling. By creating hyper-realistic visual effects, deepfakes expands the scope of creativity, while challenging traditional norms of filmmaking and demonstrating the industry's further turn towards integrating them with new technologies and values. With capabilities such as slowing down the aging of actors, resurrecting deceased actors, and creating brand new characters, this technology has the potential to completely transform the film business. These developments will not only change the storytelling style in films, but will also revolutionize the entire production process.

These innovations are constantly redefining the possibilities of storytelling and production, allowing filmmakers to push creative boundaries while addressing both technical and ethical challenges. Deepfakes represent a new technological breakthrough in film production and use machine learning algorithms trained on large amounts of data to process existing video and audio files into highly realistic synthetic media. This technology is developing rapidly and affects various aspects of filmmaking, from production technologies such as the obsolescence of actors and the reanimation of deceased performers, to narrative storytelling, where it opens up innovative creative possibilities ^[3]. Deepfakes can enhance visual authenticity, but their integration into cinema operations involves both technical and ethical considerations, which means significant changes in the way films are produced and perceived.

Research methodology

The methodology of this study is based on an integrated approach and includes an analysis of the development of deepfake technologies and artificial intelligence (AI) in the film industry in the period from 2017 to 2024. The authors of the article analyze the degree of influence of technology on the process of making films, issues of reducing production costs, and also address issues of ethics and authenticity.

The study includes the following stages:

1. Data collection for research: analysis of literature and sources on the development of artificial intelligence and deepfake technologies in the film industry during the period under review.

2. Case study: analysis of specific examples of the use of the technologies in question in famous films and advertising campaigns.

3. Impact Assessment: A study of the impact of artificial intelligence and deepfake technologies on the film industry, including creative empowerment, transformation of production methods, creative opportunities, and ethical issues.

4. Predicting the future: hypotheses and prospects for further development of artificial intelligence and Deepfake technologies and their impact on the film industry, taking into account current trends and achievements in the field of neural networks and machine learning.

5. Ethical analysis: consideration of the legal framework and ethical issues related to the use of artificial intelligence and deep forgery technologies, issues of protecting individual rights and preventing abuse.

Research on the paradigm shift (2017-2024)

Face swapping and synthetic video content began to attract attention with the advent of deepfake technology in 2017, marking an important turning point when enthusiasts and researchers began experimenting with technology and tools such as generative adversarial networks (GANS), which are often used to create manipulated media such as "face swapping," which involves swapping the object's faces. face to face with another has become a common method of creating fake images and videos ^[4]. Cinematographers soon realized the potential of this technology, and instead of relying on traditional film-making techniques where cinematographers relied heavily on computer graphics, makeup techniques, or body duplication, deepfake technology using artificial intelligence became popular to slow down the aging of actors, dubbing, or even bringing deceased actors back to life on screen. In addition, deepfake technology has significantly impacted the economic landscape of the film industry due to its low cost of production, which has inspired small studios to create high-quality visual effects that were once only available for high-budget films.

Jordan Peele's 2018 film "Barack Obama's Deepfake" would be a groundbreaking example of the use of deepfake technology in the media. The video, shot in conjunction with BuzzFeed, showed a synthesized image of Obama saying things like "Killmonger was right about his plan for world domination!", carefully constructed so that Peel's mouth movements match Obama's facial expression using sophisticated facial mapping. This particular commercial was intended to show how deepfakes can, in fact, usher in a new era of disinformation and undermine trust in digital content^[5]. Peel further stated that the public should reconsider their trust in the media, demonstrating how easy it can be to impersonate major public figures. This is a really simple public example of how to face these dangers. In 2019, Deepfake technology was used to show David Beckham presenting a malaria awareness campaign as part of the "Malaria Must Die" initiative (his voice was faked so that it seemed as if he was fluent in several languages) ^[6]. During this time, face-sharing apps also appeared, which raised greater public interest and ethical concerns in terms of concern, consent, and exploitation. In 2019, deepfake technology continued the trend of the previous year and found its application in Hollywood. Martin Scorsese's feature film "The Irishman" featured advanced VFX technology, similar to deepfakes, for aging actors such as Robert De Niro, Al Pacino, and Joe Pesci. Although they are not real deep fakes, the developers used artificial intelligence and ML to reduce the age of the actors ^[7]. Meanwhile, companies began experimenting with fakes for marketing, creating hyper-realistic fake images and episodic videos from celebrities to public figures. This has been greatly facilitated by significant improvements in machine learning algorithms, more advanced hardware, and more data for model training. In 2020, Deepfake technology has demonstrated its promise even in Hollywood. Deepfake technology was used in the movie "The Mandalorian" (2020) to demonstrate its accessibility compared to standard CGI. For example, deep fakes of Luke Skywalker created by fans have shown how artificial intelligence-driven tools can reflect or even enhance the result at much lower cost. This marked a radical change in visual effects (VFX) production, as deepfakes were cheaper and easier to produce than traditional computer graphics techniques^[8]. The documentary is another area of the film industry affected by deepfakes. This technology has changed the way real historical narratives and events are recreated and presented to the public. By allowing you to reconstruct dead people or simulate scenarios without using archival materials, deepfake expanded the creative possibilities of the genre. For example, the documentary "Roadrunner" about chef Anthony Bourdain, who died at the end of 2021, used deep sound forgery to recreate his voice in a part of the film where there were no original recordings. This sparked a debate about transparency and authenticity, with viewers initially praising the enriched narrative, bringing back inaccessible voices and faces, as well as the fine line between fact and fiction, while at the same time raising ethical questions about misinformation and trust^[9].

In 2021, the use of deepfake technology to show Bruce Willis in Russian advertising will be another step forward in using this new technology in terms of what is acceptable on the screen. The digitally recreated image of Bruce Willis began to appear in the Russian advertising campaign of MegaFon, one of the country's largest telecommunications providers ^[10]. A Russian company used deepcake's artificial neural network to transform Bruce Willie's face into that of a Russian actor of similar age and appearance. This step by both sides is a proof of concept for the limitless possibilities of using deepfake technology without actors on set. This brought Bruce Willis's attention to the news and led to a conversation about what this deepfake technology could mean from an ethical and commercial perspective. But its extension to deepfake advertising meant that other participants had their own precedent to refer to when considering comparable opportunities. This situation demonstrates a new way in which actors can potentially interact with their digital counterparts. Willis's licensing of his face for use in advertising changed the course of events and made other actors think about similar opportunities to do the same with their images. As for the actors, they could use this as a way to get new jobs for their next projects or potentially open up a completely different source of income for those who may have some physical limitations. In the Lion's Gate 2022 film Autumn, impeccable deepfake-style artificial intelligence technology was used, which replaced more than 30 cases of obscene language, in particular, F-bombs, which allowed it to receive a PG-13 rating, not R. This postproduction process included adjusting the sound and lip movements of the actors to better synchronize them with the rewritten dialogue, which allowed us to look at how artificial intelligence can effectively expand content to reach a wider audience ^[11]. Released on PG-13, the thriller sought to expand its audience and collect more box office receipts, as well as demonstrate how the growing influence of artificial intelligence in cinema can affect the regulatory and commercial requirements of the industry.

The filmmakers experimented with deep forgery to bring historical figures to life and enrich the narrative.

The use of deepfake technology in Boba Fett's Book (2022) demonstrates another achievement in the use of synthetic media in the entertainment industry. As an illustration, the series shows a young Luke Skywalker, who not only uses deepfake technology, but also uses artificial intelligence to display faces and computer graphics for a fully immersive and convincing image. It was a step forward from the use of technology in the second season of "The Mandalorian," as it served as an example of what technology can do in such a short time. The fact that deepfake technology can be implemented in television shows is a huge step forward: from an experimental method, it has turned into a practical tool to help tell stories and attract an audience. "Neighborhood Wars" (2023) is a hilarious documentary that skillfully uses deepfake technology to ridicule society, which is another example of how this technology can become a good social commentary ^[12]. Fake Neighborhood Wars (2023) is an example of how deepfake technology can be used in a comedy series. The use of deepfakes for long-form television content instead of short online videos is an important development first introduced in this ITVX series. The show demonstrates how deepfakes can be used to enhance storytelling in film and television, making it easier for viewers to perceive the visual concepts presented. Such innovations solve the problem of celebrities' inability to play certain roles, which in itself is an ethical dilemma. However, the show focuses on the issue of consent, presenting deepfakes as something more than just a new concept. This serves as a reminder that the use of synthetic materials has to do with authenticity, which requires systems to be registered for responsible use. Deepfake technology is no longer considered experimental, which indicates a reluctance to adhere to technological limitations in the name of modern culture and entertainment.

This trend culminated in the film Here (2024), which pioneered the use of real-time deepfishing technology to flawlessly age characters and transform them into believable representations, which has since meant widespread recognition that deepfishing has become just another element in any director's arsenal^[13]. Robert Zemeckis' 2024 film was shot here. Metaphysic.ai a technology similar to deepfake for the rejuvenation of actors Tom Hanks and Robin Wright. According to various reports, the film used advanced anti-aging techniques in real time based on artificial intelligence. While some news outlets call it "an unusual deepfake technology," others describe it as generative artificial intelligence that changes faces rather than the usual computer graphics techniques. Director Robert Zemeckis, who became famous for exploring the possibilities of new filmmaking technologies, presented this approach as a way to revolutionize visual storytelling. During this period of technological transformation, the global film industry, which includes Bollywood, faced the profound impact of deepfake technology, which has become very innovative, as well as its ethical challenges. Despite the fact that the introduction of synthetic media in India is at an early stage, films such as GOAT (2024) are showing promise, as reports claim that the use of AI-enhanced technologies allows for the creation of three different versions of Vijay in three different age groups. According to the Times of India, The GOAT film used modern technologies such as artificial intelligence and aging retardation, and featured a tribute to the late actor-politician Vijayakanth, who briefly appeared as M. S. Gandhi in disguise, with his voice dubbed by K. Manikandan^[14]. Similar trends can be observed with the introduction of artificial intelligence in Hollywood, which manifested itself in the effects of aging retardation used in the film “The Irishman”, which significantly changed audience expectations and transformed production processes.

Although deepfakes are a new experimental tool in Indian cinema, the technology is also being used in other activities that raise many concerns, such as the case of Rashmika Mandanna, who was the victim of a video uploaded without consent. Another example is Shah Rukh Khan's 2021 Cadbury advertisement, in which more than three hundred thousand fragments of content with his face and voice were created using deep forgery ^[15]. In addition, in 2021, they were also used to popularize highly questionable gaming platforms. Users claimed to have seen an advertisement on social media with a digitally altered Shah Rukh Khan who boasted of winning thousands of rupees in The Aviator. Created using uncoordinated audio and video, these commercials have been widely criticized for misinformation and the morality of using Deepfake technology to deceive viewers. This was noticed by the Indian government, and Prime Minister Modi raised the alarm about the possible damage this fake content could cause, and negotiations began on its regulation on social media platforms. The ethical consequences of deepfakes, from violation of consent to potential deception, are so obvious that they are fundamentally changing society, as studies analyzing the impact of synthetic media have shown.

The Impact of Artificial Intelligence and Deepfakes: Labor Struggles, Ethical Issues, and the Future of Creativity - using the example of the Hollywood strikes of 2023 and 2024.

The integration of artificial intelligence (AI) and Deepfake technology into filmmaking has sparked serious discussions about the future of creative work, ethical considerations related to authenticity and consent, and emerging production conflicts. The strike of 2023 and the terrorist attacks in Hollywood in 2024 illustrate the paradigm shift in the confrontation between human creativity and the intrusion of technological innovations focused on artificial intelligence, reflecting broader concerns about job cuts and the revision of copyright in entertainment. After a historic five-month strike in 2023, the Writers Guild of America (WGA) has entered into a tentative agreement with the studios. However, the tension in the relationship persisted, as the authors feared that scenarios created with the help of artificial intelligence could replace or further devalue their roles. Similarly, the American actors' union SAG-AFTRA went on strike over a labor dispute with the Alliance of Film and Television Producers, advocating protection against possible unauthorized use of their digital images using artificial intelligence and deep forgery technology. The strike also helped raise artists' awareness of the studios' ability to create or replicate artistic productions using this technology without providing compensation or obtaining permission from the original performers, which poses serious challenges to the principles of artistic integrity, intellectual property, and the right to publicity. The rapid growth in the number of fake images, which increased by 550% between 2019 and 2023, has further exacerbated this problem. The vast majority of these images have been examined for their negative effects ^[16]. The Hollywood strike in 2024 led to a broader discussion about the role of AI in content creation. An important issue was the misuse of AI to digitally reproduce actors' performances, which undermined traditional work models. According to Tony Vinciquerra, chairman of the board and CEO of Sony Pictures Entertainment ^[8], the consequences of these strikes were "much more serious" for the United States than originally anticipated. The likelihood of labor disputes in the future is reduced as a result of agreements that include provisions requiring higher wages and the introduction of safeguards against the use of artificial intelligence (AI) in the workplace, which contributes to the creation of a more stable work environment ^[10]. SAG-AFTRA has also advocated for the adoption of federal legislation aimed at addressing AI issues, such as the Intimate Image Deepfishing Act and the AI Labeling Act ^[11]. These legislative initiatives underscore the union's commitment to protecting its members from the unethical use of AI by promoting transparency in media production. Despite the decisions taken, the strikes have revealed serious problems related to AI in the creative industries. For example, the double strikes were completed three months before the beginning of 2024, but their effects continued to reverberate throughout the global entertainment sector ^[17]. These agreements demonstrated the fundamental role of unions in negotiating fair treatment and setting boundaries for the use of artificial intelligence in content creation. Using the power of collective bargaining, trade unions tried to curb the unrestricted and uncontrolled introduction of artificial intelligence technologies that could otherwise reduce the role of human creators ^[15]. The SAG strikes in 2023 and Hollywood strikes in 2024 highlighted the need for balanced strategies in the creative industries. These incidents highlight the importance of protecting the rights of artists and developing an ecosystem in which technological progress is displaced rather than increasing human creativity and contribution.

Research and analysis of advances in machine learning and future trends.

Advances in machine learning, improved data processing, and computing power have contributed to the development of deep forgery technologies. Over the years, these technologies will continue to evolve, creating new opportunities in cinema. deepfakes have improved quality and efficiency through the development of more complex neural networks such as transformer models. These architectures have more computing power, allow learning on more complex datasets and analyzing data in great detail, making their results realistic. Improvements in machine learning algorithms will only expand the possibilities of using synthetic media similar to live ones. One of the points of contact between deepfake technology and augmented reality (AR) and virtual reality (VR) is integration into immersive storytelling. This allows filmmakers to go beyond visual effects and use artificial intelligence to map characters in worlds so that they can be placed in stories that viewers interact with. The combination of these technologies can transform not only cinema. So, what trends are expected to determine the future of artificial intelligence-based deepfakes in films as artificial intelligence algorithms become more sophisticated, it will be much more difficult for viewers to distinguish deepfake-like footage from real ones. This feature will allow filmmakers to edit the footage on set, adding creative possibilities as well as the ability to react in real time during filming. However, the advent of more advanced and easy-to-use tools for creating deep fakes, such as DeepFaceLab and Reface, has democratized this technology and made it accessible not only to industry professionals, but also to amateurs who want to experiment with creating deep fakes ^{[18, pp. 8-13]}. As deepfake technology evolves, the need to develop ethical principles and a regulatory framework will become more and more apparent. Thus, the future of cinema based on artificial intelligence and deepfake technology is associated with unlimited new horizons, which certainly require the same ethical obligations.

Conclusion and overview of prospects

The film business has undergone many transformations since its inception, and the emergence of artificial intelligence and deepfakes between 2017 and 2024 has been a significant development due to their advanced, cost-effective, and highly accurate manipulation of human facial expression and personality. This opportunity has allowed for high-quality visual effects in filmmaking, which was previously the domain of high-budget films, but has also raised questions about consent, intellectual property, and job loss. The SAG-AFTRA strikes in Hollywood in 2023 and 2024 demonstrated these conflicts when actors demanded guarantees against unauthorized use of their digital images. Even with trade union wage agreements and the threat of automation, deepfishing technology continues to be a significant obstacle to such efforts. A good example of this is the growing interest of film producers who use Al to transform actors into younger versions of themselves, resurrect the dead, or create fully virtual characters, as reported in films such as "The Irishman" (2019) and "Here" (2024). Because such achievements reduce the cost of production and increase creativity, but at the same time, technology poses a threat to replace traditional jobs ^[19]. In the near future, in combination with augmented and virtual reality, deepfake technology can transport people to a world where reality and fiction are blurred. The risk of deepfake becoming publicly available is that it is easily abused, so work ethics and rules are very important. For example, laws such as the Intimate Image Deepfishing Act can address such issues while ensuring accountability and transparency. As technology continues to evolve, the use of artificial intelligence in film production depends on finding a compromise between innovation and the moral awareness that technological progress helps, not replaces, human imagination. Therefore, the future of cinema should lie in the limitless possibilities of human-centered development, not in technological breakthroughs.