利用AI生成知名人物形象、声音进行节目配音、视频及综艺制作的侵权与平台、模型生产商责任研究

… Infringement liability is … Deepfakes could be used to incorporate celebrities into a variety of products. Content creators have already been tapped to produce celebrity deepfakes to be …

The advent of deepfake technology has introduced new challenges to copyright law due to its ability to manipulate media content. This chapter delves into the intricate relationship between deepfakes and copyright law, examining the legal frameworks in key jurisdictions and the ethical implications for creators. Beyond the legal considerations, deepfakes also raise significant ethical concerns which need to be addressed. Deepfakes, generated by artificial intelligence, can reproduce, transform, and adapt existing copyrighted material, raising questions about authorship and originality. It is crucial to address the challenges posed by deepfakes while balancing the interests of creators, copyright holders, and the public.

Deepfake technology has rapidly evolved, enabling the seamless manipulation of facial expressions, voice tones, and lip movements, making it increasingly difficult to distinguish real from fake content. While deepfakes offer creative potential in entertainment, they also present significant challenges to intellectual property rights. This article explores the legal landscape of deepfakes in South Korea, focusing on copyright infringement, the right of publicity, and regulatory measures. It examines how deepfakes exploit copyrighted materials without authorization, raising concerns about unauthorized reproduction and derivative works. Additionally, the article discusses the misuse of deepfake technology for commercial gain and personal violations, such as deepfake pornography. South Korea has responded by enacting the AI Basic Act, inspired by international regulations, to mitigate these risks through transparency requirements and labeling obligations. The study highlights the urgent need for proactive legal frameworks and enforcement strategies to address the growing challenges posed by deepfakes.

… exists for celebrity and non-celebrity … celebrity’ and a ‘non-celebrity’ seems to be an increasingly arbitrary one.Therefore, the deepfake need not necessarily pertain to a national celebrity …

… deepfake tools makes identity itself a manipulable asset which travels across borders in seconds. Celebrities, … protection against trademark infringement. Judicial bodies have enforced …

… associated with the exploitation of celebrities, sports figures, and … My thesis is that while deepfake images of celebrities and … criminal law implications of deepfakes. For information on …

… -earning dead celebrities, … celebrities in one way or another has been a thing for decades. However, the thing that changed is the how of this business. Since not every dead celebrities’ …

… laws, but it also infringes personality rights by unauthorized … insinuation of deepfakes on the reputation of celebrities and … balance celebrity rights against malicious usage of deepfake …

… Although it must be noted that less reputable entities will use celebrities without permission.AI has made such use easy as a deepfake can be created whereas before, a photograph or …

… Deepfakes are not limited to celebrities. … including copyright and trademark infringement, but these claims are of limited effectiveness due to being conceptually removed from deepfakes' …

… , and privacy infringement, that arise when celebrities are impersonated through deepfake … Evaluate how existing frameworks respond to the unique challenges posed by deepfake …

… This chapter explores the problematic issue of voice cloning from … property recourses to tackle unauthorised voice cloning. … of the voice actors' likenesses and publicity rights, as well as …

… There is no doubt that artists known for their distinct voices are the first "victims" for AI-cloning technology, but they are not the only victims. With Al providing a nonconsensual generation …

This article explores the impact of Artificial Intelligence (AI) on the music industry, particularly focusing on the case of AI-generated covers. The emergence of AI technologies has been raising concerns not just about the originality and protection of AI-generated outputs but also about the complex input and training phase of those systems. The focus of this contribution is the latter, analysing the case of AI covers from the perspective of copyright and image rights. In the first part, an overview of the text and data mining (TDM) exception found in Article 4 of Directive 2019/790 is presented, with a primary focus on the opt-out mechanism in connection with the three-step test. Moving to the second part, the analysis delves into the complexities of voice cloning, highlighting the absence of a comprehensive European Union regime for image rights. By addressing these issues, this contribution unveiled two crucial points. First, AI models trained on various artists’ works to create and spread deepfake covers not only violate copyright but also reveal shortcomings in the TDM exception. Second, while the multifaceted image right regime may not be as exhaustive as necessary, it proves to be a viable solution against voice cloning with anticipated advancements in the future.

Generative AI has quickly expanded voice cloning’s capabilities, while also lowering barriers for using this technology. Questions about how voice cloning may impact our claim to ownership over our voices remain open, with legal standards lagging in protections for everyday people. This suggests there is a critical need for researchers to develop more frameworks to understand the complexities of voice and voice data as emerging sites where extraction and exploitation can occur. To this end, we explore voice cloning as a tension point that bridges the competing goals of how voice functions in social versus commodity frameworks. We approach this topic through two main research questions: (1) What do different paradigmatic understandings of voice reveal about the power dynamics of voice cloning? (2) How is voice characterized in legal cases and discourse around voice cloning? Using lawsuits from the United States (US) as evidence, we investigate how different voice paradigms are materialized in court decisions and argumentation. We identify key cases where voice cloning is litigated as misappropriations of identity, demonstrating the incompatibility between the value of voice as an intrinsic property of the self and the value of voice as a product. Our framework provides theoretical footholds for describing the ontological complexities of voice cloning and suggests new paths forward for researchers and policymakers.

… ), while voice cloning replicates an individual’s voice using AI models … Deepfakes and voice cloning have already been … Intellectual Property and Publicity Rights focus on ownership …

The development of Artificial Intelligence (AI) technology has had a significant impact on various fields, including social media and the entertainment industry. One innovation that has sparked controversy is voice cloning, which is the digital imitation of public figures' voices without the consent of the voice owners. This phenomenon is becoming increasingly prevalent on social media in the form of songs or audio-visual content that resembles the voices of public figures, thereby posing legal, economic, social, and ethical risks. This study aims to identify the risks faced by public figures and content uploaders due to the use of AI-based voice cloning, as well as to analyze the legal protections available under Indonesia's legal system. The research method used is a normative legal approach, combining legislative analysis and literature review. The research findings indicate that for public figures, the risks include violations of moral and economic rights, defamation, and loss of control over their personal image. Meanwhile, for content uploaders, the risks include civil and criminal liability, administrative sanctions from digital platforms, loss of credibility, and ethical consequences. Current legal protection refers to the Copyright Law, the Electronic Information and Transactions Law, the Personal Data Protection Law, the Criminal Code, and other related regulations, which are applied both preventively and repressively. However, there are no explicit regulations regarding the right of publicity and the recognition of voice as part of legally protected identity, so more adaptive regulatory updates are needed to keep pace with technological developments in the digital age.

… , concludes with urgency that India requires a standalone Right of Publicity statute.Bhat identifies the very specific inadequacy that existing Indian law protects an artist's recorded …

… publicity rights, highlighting their limitations in protecting individuals from unauthorized voice cloning. … as the proposed Federal Anti-Impersonation Right Act. Such principles can provide …

The legal issue in this study focuses on the unclear criminal law regulation of artificial intelligence-based voice cloning practices used for commercial purposes, especially in relation to Article 113 of Law Number 28 of 2014 concerning Copyright. The practice of voice cloning raises issues regarding the legal status of cloned voices as well as legal protection for the original voice owner as performers. This study aims to analyze the legal status of voice cloning from the perspective of the Copyright Law and assess whether the use of voice cloning for commercial purposes without permission can be subject to criminal sanctions based on Article 113 of Law Number 28 of 2014. The research method used is normative juridical law research. The results of the study show that the human voice is not a creation, but an object of Related Rights attached to the performer. The cloned voice cannot be categorized as an original creation that stands alone because it is formed based on existing human voice data, so legally it is more accurately understood as a form of transformation or derivative work of the object of the Related Rights. The use of voice cloning without the permission of the original voice owner for commercial purposes meets the elements of acts without rights, violation of economic rights, and commercial purposes. Therefore, juridically normative, the practice of voice cloning for commercial purposes can be qualified as a criminal act of copyright infringement and related rights and can be subject to criminal sanctions based on Article 113 of Law Number 28 of 2014 concerning Copyright, even though artificial intelligence technology has not been explicitly regulated in the law.

Generative artificial intelligence (AI) has unsettled traditional boundaries between human authorship, machine creativity, and personal identity. Systems such as large language models (LLMs), text-to-image generators, and voice-cloning technologies are trained on vast repositories of copyrighted material, extracting patterns to produce novel outputs. These outputs increasingly emulate the stylistic signatures of authors or replicate the likeness and voices of individuals without authorization. The processes of training and generation have therefore become focal points of contestation among artists, legal scholars, and policymakers, as they implicate both the integrity of copyright protection and the evolving contours of personality rights. The present paper examines the ways in which courts and policymakers are addressing the complex intersection of generative AI, copyright law, and personality rights, with particular emphasis on the protection of voice and likeness in India and in other jurisdictions. It situates recent legal and regulatory developments—such as Midler v. Ford Motor Co. in the United States, Tennessee’s ELVIS Act, and the Bombay High Court’s ruling in Arijit Singh v. Codible Ventures LLP—within broader scholarly debates concerning fair use, derivative works, originality, and the misappropriation of identity. Through a comparative lens, the analysis underscores divergent regulatory approaches: the United States’ reliance on litigation and common law doctrines; the European Union’s statutory framework prioritizing transparency, data protection, and opt-out mechanisms; and India’s evolving jurisprudence, characterized by judicial innovation through dynamic injunctions and the recognition of personality rights. The study advances a unified “Consent–Compensation–Control” (3C) framework that integrates copyright and personality rights as a coherent response to AI-driven infringements. This framework emphasizes the need for explicit consent in the use of creative works and personal attributes, equitable mechanisms of compensation for such uses, and practical legal and technological tools to maintain individual control over identity and expression. By bridging the longstanding doctrinal divide between intellectual property and personality rights, the framework aims to safeguard both creative labour and personal dignity, while simultaneously enabling innovation within responsible and clearly delineated boundaries.

… AI and the unauthorized uses of individuals' personal and bio… Deepfakes use Al to generate completely new video or audio, … broad category of rights is known as personality rights and "…

… AI-generated personalities raises questions about intellectual property rights, particularly regarding ownership of likenesses and the potential for exploitation (Vasa, 2024).The …

… voice and likeness in India and the rising trend of the judiciary in citing personality rights as … This article discusses a clash of dogma of originality and automation in AI-generated works. …

… However, the application of this understanding of personal information to AI-generated … individual’s likeness – remains untested in Australian courts. Where likenesses are constructed …

… spectrum between property rights and personality rights. On … be understood as a personality right with proprietary dimensions… In the era of digital replication and AI-generated likenesses, …

… These AI-generated songs that imitated Sun Yanzi's voice began … likeness, voices are not likenesses and cannot simply be merged with likeness when analyzing any personality rights …

Current research examines the landmark judgment of Jaikishan Kakubhai Saraf alias Jackie Shroff v. The Peppy Store & Ors. indicating the importance of developing personality rights jurisprudence in India, especially in digital content. Personality rights include both publicity rights and privacy rights. Right to control the commercial use of one's identity, i.e. publicity rights, is important for protecting personalities from unauthorized exploitation of their persona (name, image, voice, etc.). The case highlights the unauthorized use of Shroff's image for commercial gain without his consent, raising significant questions about the scope and enforcement of personality rights in India. The Court's decision to protect actor Jackie Shroff's identity from unauthorized commercial exploitation marks a pivotal moment in legal precedents safeguarding the identity of personality, especially for celebrities. This case underscores the evolving landscape of legal protections in the digital age, emphasizing the importance of rec-ognizing and upholding personality rights amidst increasing unauthorized commercial use. As a result, it stands as a landmark decision, guiding the future of personality rights and reinforcing the legal safeguards necessary to personality identity in an increasingly digital world. By dissecting this case, the paper aims to explore the im-plications for the ongoing development of personality rights in India and the growing importance of safeguarding individual identity in the digital era. Also, the paper critically assesses the current legal framework, including statutory provisions and judicial precedents. It has been compared with international standards to detect loopholes and propose improvements, in the evolving regime of personality rights in the digital era.

The article addresses the issues of legal regulation of artificial intelligence (AI) in the context of its rapid development and penetration into various spheres of life. The introduction raises the problem of uncertainty in the legal environment regarding content created using AI, with a particular focus on copyright issues and the possibility of legislative recognition of AI as a subject of law. Market statistics analysis shows the growth of the global AI market and underscores the importance of developing legislation governing authorship and intellectual rights to prevent potential legal disputes and protect personal data. Special attention is given to the differences in approaches to the legal status of AI worldwide, including the USA, the UK, the European Union, China, and Russia, as well as initiatives by international organizations such as UNESCO and the World Intellectual Property Organization. The research methodology is based on comparative legal analysis, examination of regulatory acts, and expert evaluation, which allowed for identifying common features and significant differences in AI regulation across various jurisdictions. The article also explores key ethical issues related to the use of AI, including personal data protection and preventing data leaks. The research aims to propose possible solutions and adaptations of legislation considering the rapid development of AI technologies. It will be useful not only for lawyers and intellectual property specialists but also for a wide range of readers interested in modern technologies and their legal aspects.   Received: 22 July 2024 | Revised: 25 October 2024 | Accepted: 11 February 2025   Conflicts of Interest The author declares that she has no conflicts of interest to this work.   Data Availability Statement Data sharing is not applicable to this article as no new data were created or analyzed in this study.   Author Contribution Statement Anna Pokrovskaya: Conceptualization, Formal analysis, Writing – original draft, Writing – review & editing, Visualization, Funding acquisition.

… not account for synthetic media technologies and therefore … (Al) and synthetic media technologies have woven a tapestry … the doctrine of intermediary liability with a constitutionally …

… of synthetic content, not just initial uploads, while clarifying that intermediary liability applies … without constraining lawful innovation, creativity, or the responsible use of synthetic content. …

With the explosion of computer technology, vastly more and more varied types of data related to individuals are being disseminated online, often without their consent. While intermediary publishers are not the initial and immediate cause of this, they generally play a contributory role and engage in further (semi-)autonomous processing such as organizing or promoting content. Current case law rather haphazardly recognizes intermediary publishers to be data protection ‘controllers’ and/or protected by the intermediary ‘host’ shield, while also acknowledging the engagement of general human rights law. Seeking to synthetically balance the competing purposes which underlie these three legal frameworks, this article argues that greater responsibility should flow from more autonomous control but that some shielding is still necessary for all intermediary publishers. Conceptually it is argued that such a synthetic approach leads to intermediary publishers being grouped into three increasingly autonomous categories—‘processor hosts’, ‘controller hosts’ and ‘independent intermediaries’—which should be subject to a successively greater ambit of responsibility accordingly. Detailed elaboration of the resulting duties must also take account of the seriousness of the potential interference with competing rights and, in this regard, should give weight to the divergent resource capacity of otherwise similarly situated actors.

Deepfake and other synthetic-media harms create a civil-liability problem that ordinary tort doctrine does not easily resolve: harmful content may be generated, amplified, monetised, and redistributed through a chain of actors in which no single participant controls the whole causal process. The objective of this article is to develop a layered civil-liability framework for that problem. It examines the Saudi and Jordanian civil-liability regimes as the principal doctrinal focus, while using selected EU, US, UAE, and Chinese materials as comparative reference points. The article remains private-law centred. It asks how civil-liability doctrine can respond when reputational, identity-based, corporate, and non-material harm is produced through the combined conduct of generative-model developers, prompting users, online platforms, and secondary distributors. The analysis identifies three pressure points: the natural-person wording of Article 138(2) of the Saudi Civil Transactions Law, which narrows moral-harm protection in relation to corporate and institutional injury; the persistence of a single-wrongdoer model in Saudi and Jordanian doctrine, despite the layered actor structure of deepfake production; and the evidential asymmetry created by opaque algorithmic causation. The article’s main contribution is a calibrated model of layered civil liability: clearer protection for juridical-person reputation, a custody-based rule for algorithmic systems, cautious burden shifting where relevant information is controlled by platforms or developers, stronger private-law links to data-protection regimes, and targeted transparency duties for synthetic-media systems. The contribution is not to replace civil law with AI regulation, but to show how both fields must work together if deepfake-induced harm is to be remedied in a legally coherent way.

… and/or protected by the intermediary ʻhostʼ shield, whilst also acknowledging … intermediary publishers. Conceptually it is argued that such a synthetic approach leads to intermediary …

… While generative art and synthetic media were among the first of the generative AI content to become popular, it was the release of ChatGPT that transformed the growth of generative AI …

… Part Ii defines terms and describes the state of intermediary liability law prior to Congress's action and then recounts the passage of the statutory immunity provision, including its broad …

ABSTRACT New approaches to the legal duties of Internet intermediaries are emerging. Current critiques of technology companies in what is said to be a ‘techlash’ overlap with the proposing of new models of liability and responsibilities. Do these shifts in attitude, and the associated set of new ideas, mean that legislative bodies might be more willing, today, to revisit the balance struck in the late 1990s? Changes and challenges to the general provisions applicable to intermediaries, and the introduction of standalone provisions in specific sectors (such as audiovisual media regulation and copyright) are discussed; emphasis is placed on the proliferation of ‘voluntary’ measures (e.g. on illegal content and on disinformation), which provide evidence of changing attitudes. Further arguments include the overlap between available causes of action in relation to Internet communications (e.g. data protection and harassment law), with implications for jurisdiction, remedies, and other matters, and the attractiveness of alternative approaches, including the cross-cutting control of ‘harmful digital communications’ in New Zealand, and proposals to apply specific regulatory regimes, influenced by financial regulation and other fields, to online material. The UK government’s recent ideas regarding a possible ‘duty of care’ for certain intermediaries are assessed in the context of these developments.

… the communication containing synthetic media" a disclosure that the media was "generated … includes any synthetic media shall be required to disclose the use of such synthetic media. …

The legal landscape surrounding Internet intermediaries-such as e-commerce platforms, social media, and ISPs-centers on their liability for third-party actions. Such liability can be framed as direct liability, for instance, for defective products through strict product liability, and as indirect liability, for instance, in actions for defamation and copyright infringement. The general approach however has been to shield liabilities from liability. Legislation such as the US Communications Decency Act, the US Digital Millennium Copyright Act , the EU Electronic Commerce Directive generally exempt intermediaries from liability, provided the various procedural mechanisms in such legislation are observed. However, it is noteworthy that the EU's Copyright Directive 2019 and EU Digital Services Act 2020 mandate proactive measures, including implementing content filtering and conducting due diligence risk assessments. As the legal scrutiny over Internet intermediaries intensifies globally, balancing accountability with intermediary immunity remains critical to sustaining innovation and growth on the Internet.

The rise in the integration of Artificial Intelligence (AI) into media productions for example remastering, dubbing and content adaptation has raised significant legal and ethical concerns. Central to these issues are the rights to adaptation under copyright law, which grant creators exclusive authority over derivative works and transformations of their original creations. These rights are often challenged by AI-driven processes like audiovisual quality enhancement, synthetic voice generation, cultural localization raising questions about ownership, fair use, and copyright compliance. This research proposal will therefore examine the legal effects of AI enhanced adaptations through a thorough evaluation of relevant case law, jurisdictional differences and market forces so as to strike a balance between innovation and maintaining respect for artists' rights.

This article offers an academic reflection on creativity in dubbing, considering the emergence and growing presence of AI-dubs. It explores the concept of creativity as a possible key differentiator between human agency and machine-generated content (Brandt 2023; Grassini and Koivisto 2024; Runco 2023). To this end, the article first examines creativity through a cross-disciplinary lens, establishing a broader conceptual theoretical framework and context for its definition and meaning. It then explores the creative practices associated with traditional studio-generated dubbing from both industry and academic perspectives, drawing insights from other fields that engage in constrained creative processes (Baetens 2010; De Geest and Goris 2010; Krauth 2016; Wang 2009). Finally, it rethinks traditional creative practices in the context of AI-generated dubbed content, briefly summarizing preliminary research findings that compare studio-dubs and AI-dubs. The discussion focuses on how AI-dubbing technologies challenge established boundaries and norms, reshaping creative practices and the roles of dubbing practitioners — especially translators — and potentially influencing viewer perception.

… on how this work has been extended to also allow for massive machine dubbing of MediaUPV… On the contrary, being a derivative work of an educational video owned by a lecturer, (raw) …

Objective: With the profound transformation and upgrading brought by the new wave of intelligent media and digital empowerment, artificial intelligence is reshaping various aspects of art, including creation, distribution platforms, dissemination channels, and feedback effects. Firstly, personalized, innovative, and diversified artistic creation processes, along with intelligent, customized, and precise artistic dissemination processes, are cultivating "new aesthetic hotspots" in the era of intelligence. Secondly, amidst the landscape of technological change, ideological and ethical issues have become increasingly prominent, revealing the risks and obstacles of the "Artificial Paradise". Methods: Using the literature review method, conduct a comprehensive discussion on the new landscape of Film and Television Arts development. Conclusions: In response to these new changes and trends, the development of film and television arts in the era of artificial intelligence requires the anchoring of new policies and measures.

Abstract Digital interactive accessibility services must enable human interaction with the media content beyond consumption, ensuring that people can perceive, understand, navigate, and interact with the content and contribute to it. All these new types of interactions are also a field of study in audiovisual translation (AVT) and media accessibility (MA), as translation is a form of human–computer interaction (HCI) (O’Brien, Translation Spaces 1:101–122, 2012). Therefore, the role of technology in AVT/MA is a key aspect not only in the process of creation, but also for distribution, delivery, and consumption. This tendency is expected to grow in the coming years as “technology is also the basis of tools to translate or adapt content and tools to consume content” (Matamala in Accessibilitat i traducció audiovisual, 2019). The progressive transition from Web 2.0 to Web 3.0, with the irruption of recent technologies such as blockchain and artificial intelligence (AI), is opening up innovative forms of communication and interaction for users in the digital world. Yet, challenges in relation to intellectual property rights (IPR) management in AVT/MA remain a major concern (Orero et al., 2023; Serrat-Roozen & Oncins, 2023). The following chapter presents the results of a series of focus groups held with professionals and researchers from the different fields of the AVT and MA in the frame of the European project MediaVerse. This three-year project was aimed at designing and testing a framework to allow professionals and laymen to publish multimedia content that may be easily shared. Results presented highlight user’s needs and expectations from AVT/MA professionals in relation to copyright management, through the use of blockchain technology to protect and recognise IPR for professionals in these fields.

This case study focuses on the application of Artificial Intelligence (AI) in the advertisements of Online Travel Agency (OTA). With the development of AI, companies have more tools to create advertising. Deepfake, as a technology powered by Artificial Intelligence, could be used to manipulate the content in advertising. As a result, it is important to review the effect and users’ attitudes toward AI-advertising. This case study reviews the Trivago AI advertising and discusses the associated theory in the AI-human reaction. The possible solutions are also discussed in the case study.

Generative artificial intelligence (AI) technology is increasingly used across various fields. It is particularly prominent in creative industries such as television, film, advertising, gaming, and entertainment. Despite its creative application, AI technology raises certain legal and ethical issues at a global level. This research investigates whether audio visual works generated through generative AI technology are protected under international copyright law. The qualitative method was used for this study, and the primary data included the Sri Lanka Intellectual Property Act No. 36 of 2003, the Indian Copyright Act (1957), and the US Copyright Act (1976) which were studied in a comparative manner, and the main case study was ‘Suryast AI Image’. The data were analyzed under the thematic analysis method, and how authorship, ownership, exclusive right, and moral right are addressed within the copyright law of the three countries. This research confirmed that Sri Lanka, the United States and India currently lack legal provisions to protect AI-generated artwork under their existing copyright laws. The lack of legal consensus on who is the author of an AI-generated work is the main problem with granting copyright to an AI-generated work. Establishing a clear definition of authorship AI-generated work, protection for AI-generated audio-visual works can potentially be provided either through existing copyright laws or via new legislative amendments. Since AI has created the next era of art, the protection of a copyright legal framework for it has become urgent.

… The rise of social media has provided a platform for the use of celebrities… celebrities and sportspeople to establish, the real hurdle is posed by misrepresentation. Unlike the strict liability …

The development of deepfakes began in 2017, when a software developer on the Reddit online platform began posting his creations in which he swapped the faces of Hollywood celebrities onto the faces of adult film artists, while in 2018, the comedic actor Jordan Peele posted a deepfake video of former U.S. President Obama insulting former U.S. President Trump and warning of the dangers of deepfake media. With the viral use of deepfakes by 2019, the U.S. House Intelligence Committee began hearings on the potential threats to U.S. security posed by deepfakes. Unfortunately, deepfakes have become even more sophisticated and difficult to detect. With easy accessibility to the applications of deepfakes, its usage has increased drastically over the last five years. Deepfakes are now designed to harass, intimidate, degrade, and threaten people and often leads to the creation and dissemination of misinformation as well as creating confusion about important state and non-state issues. A deepfake may also breach IP rights e.g., by unlawfully exploiting a specific line, trademark or label. Furthermore, deepfakes may cause more severe problems such as violation of the human rights, right of privacy, personal data protection rights apart from the copyright infringements. While just a few governments have approved AI regulations, the majority have not due to concerns around the freedom of speech. And while most online platforms such as YouTube have implemented a number of legal mechanisms to control the content posted on their platforms, it remains a time consuming and costly affair. A major challenge is that deep fakes often remain indetectable by the unaided human eye, which lead to the development by governments and private platform to develop deep-fake detecting technologies and regulations around their usage. This paper seeks to discuss the legal and ethical implications and responsibilities of the use of  deepfake technologies as well as to highlight the various social and legal challenges which both regulators and the society face while considering the potential role of online content dissemination platforms and governments in addressing deep fakes.

… The impact of deepfake technology is not restricted to celebrities; even wellmeaning … It is therefore doubtful that a deepfake victim could hold a platform or media company liable under …

… sword imposing greater platform liability. Grounding the platform liability framework more … When a deepfake video of a celebrity engaging in explicit sexual acts goes viral on social …

Deepfake technology is spinning out of control. Recent advancements allow users to quickly, easily, and anonymously create fake yet highly realistic images and videos featuring real people. While this technology has potential benefits, it is widely used nefariously to create pornographic images and videos of young girls. Ninety-eight percent of all deepfake videos online are pornographic deepfakes. The number of deepfake pornography videos has gone up 464% since 2019 and is expected to continue increasing as technology develops and becomes even more accessible. Scholars have taken note of this frightening trend, seeking to establish some form of liability for the harms generated by sexual deepfakes. This research has focused on two key concepts: user accountability and platform accountability. Yet scholars also show that these two theories of liability are bound to fail. Identifying users who create deepfakes is extremely challenging at best, and usually unhelpful, as they are typically judgment proof. And platforms, where deepfakes are widely shared, are protected from liability by Section 230 of the Communications Decency Act. Against this backdrop, this Article offers a third solution by proposing a novel approach to deepfake liability hitherto not discussed in the literature: common law manufacturer accountability. This proposal focuses on the companies developing generative Artificial Intelligence (“AI”) tools and offers a natural response to the deepfake crisis with a high likelihood of success. The companies developing generative AI models are manufacturing dangerous and unsafe tools, with no effective safeguards to minimize associated harms. These companies can therefore be found liable for the resulting harm based on familiar concepts of products liability law and design defect. These companies are not protected under Section 230, are incredibly profitable and powerful, and have a crucial impact on the AI ecosystem. We show that manufacturer liability will provide much-needed protection to the victims of sexual deepfakes. We also show that, by inducing AI companies to include basic safety measures in their products, our proposal will have positive spillover effects on political deepfakes, where the misuse of AI technology currently threatens to destabilize democratic order and to expose democratic processes to dangerous foreign manipulations. Our proposal addresses a critical oversight in current research and regulatory discourse aimed at tackling the deepfake crisis. The widespread integration of generative AI into user-friendly interfaces has made this powerful technology accessible even to young children, who can now use it to inflict significant harm with minimal technical skills. This urgent reality necessitates a new approach— one that reassigns liability and responsibility to the companies that develop and deploy the technology enabling deepfakes.

The rapid development of Artificial Intelligence (AI) has introduced deepfake technology, which is capable of manipulating audiovisual data with high realism. While this technology offers creative utility, its misuse for synthesizing non-consensual pornography, political disinformation, and character assassination severely violates an individual’s right to reputation. This study examines the civil and criminal liabilities of AI deepfake platform providers under the Indonesian legal framework, specifically Law No. 1/2024 (Second Amendment to the ITE Law) and Law No. 27/2022 on Personal Data Protection (UU PDP). Using a socio-legal research method, this study analyzes how textual legal norms interact with the digital behavior of platform providers and the psychological and social vulnerabilities of victims. The results indicate a significant normative gap: current regulations primarily target the individual creators or distributors of deepfake content, while platform providers escape liability by hiding behind the "safe harbor" doctrine. Sociologically, this legal vacuum fosters a culture of impunity and exacerbates digital gender-based violence. This paper argues for a paradigm shift from passive moderation to active algorithmic accountability. Platform providers must be held civilly liable under Article 1365 of the Indonesian Civil Code for systemic negligence, and criminally liable under the corporate fault doctrine for failing to implement mandatory digital watermarking and proactive content filtration systems.

The complexity and emergent autonomy of Generative AI systems introduce challenges in predictability and legal compliance. This paper analyses some of the legal and regulatory implications of such challenges in the European Union context, focusing on four areas: liability, privacy, intellectual property, and cybersecurity. It examines the adequacy of the existing and proposed EU legislation, including the Artificial Intelligence Act (AIA), in addressing the challenges posed by Generative AI in general and LLMs in particular. The paper identifies potential gaps and shortcomings in the EU legislative framework and proposes recommendations to ensure the safe and compliant deployment of generative models.

Large generative AI models (LGAIMs), such as ChatGPT, GPT-4 or Stable Diffusion, are rapidly transforming the way we communicate, illustrate, and create. However, AI regulation, in the EU and beyond, has primarily focused on conventional AI models, not LGAIMs. This paper will situate these new generative models in the current debate on trustworthy AI regulation, and ask how the law can be tailored to their capabilities. After laying technical foundations, the legal part of the paper proceeds in four steps, covering (1) direct regulation, (2) data protection, (3) content moderation, and (4) policy proposals. It suggests a novel terminology to capture the AI value chain in LGAIM settings by differentiating between LGAIM developers, deployers, professional and non-professional users, as well as recipients of LGAIM output. We tailor regulatory duties to these different actors along the value chain and suggest strategies to ensure that LGAIMs are trustworthy and deployed for the benefit of society at large. Rules in the AI Act and other direct regulation must match the specificities of pre-trained models. The paper argues for three layers of obligations concerning LGAIMs (minimum standards for all LGAIMs; high-risk obligations for high-risk use cases; collaborations along the AI value chain). In general, regulation should focus on concrete high-risk applications, and not the pre-trained model itself, and should include (i) obligations regarding transparency and (ii) risk management. Non-discrimination provisions (iii) may, however, apply to LGAIM developers. Lastly, (iv) the core of the DSA's content moderation rules should be expanded to cover LGAIMs. This includes notice and action mechanisms, and trusted flaggers.

… Generative AI is defined in this research context as a branch of machine learning relying on foundation models … provider, rendering them the legal custodian liable for inherent defects.…

… As the business models of generative artificial intelligence continue to evolve, the … liability on the providers of generative artificial intelligence technology. To clarify their tortious liability, it …

… to smart-home technologies and generative AI. It is for this … Providers of generative AI systems are required to comply … of liability for open-source foundation model providers would …

In Europe, the governance discourse surrounding artificial intelligence (AI) has been predominantly centred on the AI Act, with a proliferation of books, certification courses, and discussions emerging even before its adoption. This narrow focus has overshadowed other crucial regulatory interventions that promise to fundamentally shape AI. This article highlights the proposed EU AI liability directive (AILD), the first attempt to harmonize general tort law in response to AI-related threats, addressing critical issues such as evidence discovery and causal links. As AI risks proliferate, this article argues for the necessity of a responsive system to adequately address AI harms as they arise. AI safety and responsible AI, central themes in current regulatory discussions, must be prioritized, with ex-post liability in tort playing a crucial role in achieving these objectives. This is particularly pertinent as AI systems become more autonomous and unpredictable, rendering the ex-ante risk assessments mandated by the AI Act insufficient. The AILD’s focus on fault and its limited scope is also inadequate. The proposed easing of the burden of proof for victims of AI, through enhanced discovery rules and presumptions of causal links, is insufficient in a context where Large Language Models exhibit unpredictable behaviours and humans increasingly rely on autonomous agents for complex tasks. Moreover, the AILD’s reliance on the concept of risk, inherited from the AI Act, is misplaced, as tort liability intervenes after the risk has materialized. However, the inherent risks in AI systems could justify EU harmonization of AI torts in the direction of strict liability. Bridging the liability gap will enhance AI safety and responsibility, better protect individuals from AI harms, and ensure that tort law remains a vital regulatory tool.

… the future of deepfakes. Considering the centrality of personality rights to deepfakes, this … The fact that images, voice, data, and other audio-visual contents are involved means that in …

<jats:p>.</jats:p>

… If there is one development that has marked a shift in audiovisual production since the … how deepfakes intersect with broadcast regulation rather than traditional copyright or publicity …

… in the form of "deepfakes," synthetic audiovisual media that appear authentic and … deepfakes; and 2) a federal right of publicity with a presumption against the inheritability of deepfake …

… copyrighted work, thereby resolving asymmetries between expansive copyright protection and uneven publicity … argues that anchoring deepfake regulation in copyright law is …

<jats:p>.</jats:p>

This article examines the legal, regulatory and societal challenges posed by deepfake technology, situating its analysis within a comparative framework spanning the European Union, United States and China. It explores the multifaceted harms of deepfakes – from non-consensual pornography and political disinformation to financial fraud and identity manipulation – and analyses their rapid dissemination through online ecosystems that undermine both individual dignity and democratic trust. The study assesses the EU Artificial Intelligence Act, highlighting its transparency-based approach, definitional boundaries and classification of deepfake systems as ‘limited-risk’, while identifying contexts that may warrant high-risk or prohibited status. It underscores the limitations of transparency obligations in addressing malicious actors, cross-border disinformation and intimate image abuse, and examines the complementary roles of the Digital Services Act, the General Data Protection Regulation and the EU Directive on combating violence against women in regulating different stages of the deepfake lifecycle. Ultimately, the article argues for a multi-layered, adaptive governance model that reconciles the protection of rights, dignity and democratic integrity with the preservation of legitimate innovation in AI-driven creativity and communication.

The relevance of this study stems from the rapid development of generative technologies that enable the creation of heavily modified or fully synthesised content using artificial intelligence, particularly deepfakes. Such content not only creates an illusion of authenticity but also poses a threat to the protection of intellectual property rights and personal non-property rights, giving rise to significant legal challenges in the digital environment. The aim of the article was to formulate and justify civil-law approaches to the regulation of deepfake content usage in the context of copyright and personal data protection, considering the challenges of society’s digital transformation. The study employed methods of systems analysis, legal-logical generalisation, formal legal method, and comparative legal research, considering international norms and doctrinal sources. It was established that current Ukrainian legislation does not define deepfake content as a separate legal category, and existing legal mechanisms are fragmented and do not cover all aspects of responsibility for its creation and distribution. Gaps were identified in the regulation of derivative digital works, the protection of biometric features of individuals, and the procedures for identifying violators in the context of automated content generation. The study proved that without proper regulatory response, deepfake technologies may be used as tools of manipulation, identity forgery, and digital defamation. The research emphasised the need for an interdisciplinary approach that integrates legal, technical, and ethical aspects of deepfake regulation. Special attention was given to legal liability in cases of automated content creation without direct human authorship. The importance of digital transparency and informed consent was highlighted as key principles of legal regulation. The results of the study can be used to improve national legislation and to develop international legal mechanisms in the field of artificial intelligence

… available in trademark, copyright, patent, and trade secret law provide useful models for developing analogous penalties for deepfake-related identity theft under the right of publicity. …

… With deepfakes and other synthetic media technologies now capable of producing hyper … Copyright, the primary legal framework for creative ownership, offers limited protection for …