Ethical and legal implications of using deepfakes in commercials: analysis of the Volkswagen Brazil case and the possibilities of blockchain

Ethical and legal implications of using deepfakes in commercials: analysis of the Volkswagen Brazil case and the possibilities of blockchain | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Ethical and legal implications of using deepfakes in commercials: analysis of the Volkswagen Brazil case and the possibilities of blockchain Diocélio Dornela Goulart, Rodrigo Baroni de Carvalho, Ramon Silva Leite, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8118450/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The increasingly sophisticated quality of deepfakes makes it difficult for the public to distinguish between real and manipulated content, raising serious ethical and legal questions. This paper explores the ethical and legal complexities of using deepfake technology by analyzing the Volkswagen Brazil advertising case, which recreated the image of the late 80’s Brazilian popular music singer Elis Regina. Based on the decision of the Brazilian National Council for Advertising Self-Regulation (CONAR), the paper discusses the challenges advertisers face when integrating advanced technologies into marketing activities, while accounting for the perspectives of different stakeholders, including heirs of the artists, on consent for the use of posthumous images. The findings indicate a pressing need for more straightforward, comprehensive ethical and legal guidelines that address both explicit consent and the social and cultural impacts of AI use in advertising, especially for vulnerable audiences. The study underscores the importance of coordinated regulation adapted to the complexities of emerging technologies such as deepfakes, highlighting the nascent stage of current laws and regulations and the need for an approach that encompasses legal, social, and cultural aspects. As a potential technological solution, this study highlights the use of blockchain technology to address these issues. Artificial Intelligence (AI) deepfake Advertising and Propaganda blockchain digital ethics digital marketing Figures Figure 1 Figure 2 Figure 3 1. INTRODUCTION Deepfake is an AI-based technology that allows for the creation or manipulation of video or audio content in a highly realistic way. Using machine learning algorithms, deepfakes can replace one person's face with another's, alter facial expressions, or even create videos of people saying or doing things they never actually did (George & George, 2023). The increasingly sophisticated quality of these simulations makes it difficult for the public to distinguish between real and manipulated content, raising serious ethical and legal questions (Chesney & Citron, 2019; Rini, 2020; Rana et al., 2022). The issue of deepfakes has generated significant interest among marketing professionals and the general population. Despite their relevance to marketing theory and practice, deepfakes—and the opportunities for benefit or harm they provide—are poorly understood (Whittaker, Letheren & Mulcahy, 2021). Mazis (1997), Petty (2005), and Celsi (2023) have already highlighted the need for effective public policies to regulate marketing activities, and this is even more pertinent in the context of emerging technologies such as deepfakes. In the first episode, "Be Right Back," of the second season of the Netflix series Black Mirror, produced in 2013, the themes of grief, digital inheritance, and the ethical use of Artificial Intelligence (AI) and robotics are explored. Martha, the protagonist, loses her boyfriend Ash in a car accident and is enrolled by a friend in a service that creates a digital replica of Ash using AI. Initially reluctant, Martha interacts more with the replica, which evolves into a robotic form. The story highlights the limitations of technology: the replica, although similar, lacks some of Ash's characteristics and memories, revealing a less assertive side of the digitization of human identity (InternetLab, 2017; Lacerda & Mattos, 2020). On July 4, 2023, Volkswagen Brazil launched a commercial to celebrate the brand's 70th anniversary in the country, featuring a virtual simulation of the legendary Brazilian popular music singer Elis Regina. The ad recreates an interaction between Elis Regina (who died tragically in 1982) and her daughter, Maria Rita, using Artificial Intelligence (AI). In the commercial, while Maria Rita drives the 100% electric version of the Volkswagen Kombi, Elis drives the classic model, providing a moving duet between mother and daughter (The Drum, 2023; Shoot Online, 2023; DataSciConnect, 2023). As seen in the Volkswagen commercial with Elis Regina and in the Black Mirror episode "Be Right Back," the digital recreation of deceased individuals raises ethical and emotional questions about AI. Both examples show how technology can revisit past relationships and highlight the debate about the emotional and ethical implications of digital recreation using AI and deepfakes. The advancement of deepfakes has brought serious challenges, as exemplified in the 2024 elections in Brazil. The Brazilian Superior Electoral Court (TSE) faced the spread of manipulated videos that harmed candidates, such as São Paulo mayoral candidate Tabata Amaral, who was the target of false content of a sexual nature, and responded by regulating the use of AI in election campaigns (Superior Electoral Court, 2024; UOL Notícias, 2024). In addition, Starlink faced fraudulent ads that used deepfakes to clone the voices of politicians, such as the Brazilian federal deputy Nikolas Ferreira, promoting false Internet offers (Boatos.org, 2024). These cases highlight the need for regulatory, legal, and technological measures to mitigate the risks posed by malicious AI use and protect public and commercial integrity. On July 10, 2023, the Brazilian National Council for Advertising Self-Regulation (CONAR) initiated an ethical proceeding to investigate Volkswagen's advertisement that used deepfake and AI techniques to recreate singer Elis Regina. The proceeding was opened following consumer complaints questioning the ethics and the potential for confusion between fiction and reality, especially for children and adolescents. The creation of a commercial that uses AI to recreate a deceased person, as in the case of Elis Regina, raises ethical and social issues beyond simple product or brand promotion. This practice raises crucial questions about consent, image rights, and cultural impact, which need to be rigorously examined (Harris & Carman, 1983). In addition, there is a significant externality related to the potential for confusion between fiction and reality, especially among more vulnerable audiences such as children, adolescents, and the elderly (Turkle, 1997). This confusion can have psychological and social implications, such as distorting perceptions of the recreated person and their legacy. Hence, the use of advanced technologies in marketing strategies requires in-depth ethical debate and possibly the intervention of public policies to regulate such practices (Mittelstadt et al., 2016; Naffi et al., 2023; Kumar et al., 2024). The importance of effective public policies to control marketing activities, especially regarding emerging technologies, has been discussed in previous studies (Mazis et al., 1997; Petty & Celsi, 2023; Kshetri et al., 2023) that highlighted the need for careful regulation to protect consumers' and society's interests amid technological advances. The central issue of this study arises from the gap between technological advancement and the ethical and legal regulations that can mitigate the negative impacts of these practices. The recreation of deceased individuals through AI, especially for commercial purposes, raises important questions: to what extent is it ethical to recreate someone without their explicit consent digitally? What are the social and cultural consequences of this practice? How can we ensure that the public and especially vulnerable groups such as children and adolescents understand the fiction involved? How can we regulate such practices to balance technological advances and respect for human rights? This paper has potential contributions in the domains of digital marketing, political marketing and ethical challenges. In the social sphere, it seeks to deepen understanding of the emotional and cultural impacts of using deepfakes and AI to recreate public figures, thereby raising awareness of the issue. Academically and theoretically, the paper addresses the interdisciplinary debate involving technology, ethics, and communication, offering a critical overview of the externalities of digital marketing and proposing innovative solutions based on emerging technologies, such as blockchain. From a marketing perspective, the Volkswagen case analysis serves as a concrete example for brands that wish to explore advanced technologies responsibly and ethically, offering practical guidelines to avoid negative impacts on public perception. For society, the paper contributes by proposing ways to develop public policies that regulate the use of AI and deepfakes, drawing on experiences such as the AI Act approved by the European Parliament, thereby strengthening the legal framework necessary to address the implications of these technologies. By investigating the externalities of digital marketing and proposing technological and regulatory solutions, this paper not only analyzes the Volkswagen case but also contributes to the understanding of the role of ethics and innovation in a scenario increasingly dominated by AI. Thus, the paper seeks to foster a qualified, interdisciplinary debate on the limits and possibilities of deepfakes in commercial contexts, promoting a balance between technological advancement and social responsibility. Beyond regulatory issues, this paper advocates using blockchain technology as a potential solution to the problem. Dixon (2024) and Seneviratne (2022) highlight the potential of blockchain technology in combating the spread of deepfakes. Dixon (2024) emphasizes that the immutable and transparent record of blockchain is fundamental to the authentication and integrity of multimedia content, facilitating source identification and combating fake videos. Additionally, Seneviratne (2022) outlines the importance of a decentralized and cryptographic system provided by blockchain. When combined with AI techniques, blockchain could verify the authenticity of content and create an integrated approach to tackling false information and deepfakes across different platforms. Following the perspective of Motta (2022), this paper proposes solutions with a focus on extrapolation, identifying existing approaches to help reduce the impact of the emerging problem of deepfakes and AI-related spread. The focus for the solution lies in deepening the regulatory debate and using emerging technologies, such as blockchain. Beyond this introduction, the paper is structured as follows: background discussing the subjects of deepfake and its regulations and impacts on vulnerable populations; a methodology section detailing the literature review procedures and the documentary analysis; a data analysis section presenting the case of the Brazilian singer deepfake commercial; discussion about the possibilities of using blockchain to mitigate these issues, and, the conclusions. 2. BACKGROUND The Background section is structured into interconnected subsections that progress from the technological and ethical implications of AI and deepfakes (2.1) to the regulatory frameworks addressing these technologies (2.2) and finally to the social impacts on vulnerable populations (2.3). This organization establishes a coherent analytical basis that integrates technological, legal, and societal dimensions, supporting a comprehensive understanding of the deepfake phenomenon within contemporary debates on ethics, governance, and digital transformation. 2.1 Impacts of AI and deepfakes Artificial Intelligence (AI) and deepfake technology have their roots in progressive advances in machine learning and digital manipulation. Initially, deep learning algorithms were applied to recognize patterns in massive datasets (Goodfellow et al., 2014 ), enabling the creation of highly realistic content. Deepfakes, specifically, began to be widely discussed in 2017, when generative adversarial networks were used to create videos and images that even the most experienced analysts could not distinguish from real content (Kietzmann et al., 2021 ). Henceforth, technological advances have expanded their applications, from interactive entertainment to malicious uses, such as the dissemination of misinformation or political manipulation (Chesney & Citron, 2019 ). These technological advances have brought benefits in fields such as marketing and education, but they have also raised ethical challenges. For example, the use of deepfakes in advertising, such as recreating singer Elis Regina in a Volkswagen commercial, highlights how technology can affect both consumer engagement and trust in digital content (George & George, 2023 ). The ethical analysis of the use of Artificial Intelligence (AI) to recreate deceased people is a complex issue that begins with the dilemma of consent. since the recreated person cannot give their own consent (Faden, Beauchamp, and Kass, 2014 ). This matter raises complex questions about who has the right to give this consent in their place, whether it be the family, the heirs, or a legal entity. This issue extends to image rights, which are protected by law. However, laws can vary from country to country and may not be up to date to address the specific challenges presented by AI (Lessig, 2006 ). This context creates a legal vacuum that could allow for the misuse of a person's image, leading to possible legal action. In addition to legal issues, there are also considerations regarding the social and cultural impact of recreating public figures. Post (2000) and Naffi et al. ( 2023 ) argue that this can have a significant impact on society, possibly distorting or trivializing the legacy of the recreated person. This is especially concerning when the figure in question has an important cultural or historical legacy. The issue becomes even more complex when considering the risk of confusion between fiction and reality. Turkle ( 1997 ) points out that this risk is exceptionally high among children, adolescents, and the elderly, who may not fully understand that the person has been digitally recreated. Naffi et al. ( 2023 ) reiterate that this can have psychological and social implications, including distorting perceptions of the recreated person and their legacy. The increasing simplicity involved in creating convincing deepfakes, combined with our increasingly digitally documented lives, may increase the potential for deepfakes to be used for malicious purposes, such as misinformation or fake news, blackmail, intimidation, sabotage, harassment, defamation, revenge pornography, identity theft, and bullying (Whittaker et al., 2021 ). On the other hand, through deepfakes, it is possible to create dynamic advertisements for individual consumers—the ultimate in personalization. Evidence suggests that advertising personalization is positively related to ad persuasiveness and perceived benefit (Ham, 2017 ). Thus, increased personalization potential may lead to greater desired consumer (re)action (Kietzmanna, Millsb & Planggerc, 2021). Given this complex context, Mittelstadt et al. ( 2016 ) advocate the creation of strict guidelines and regulations for the use of AI to recreate deceased people. These guidelines could include explicit consent requirements, clear disclosure that the person has been digitally recreated, and possibly restrictions on how and where these recreations can be used. In addition to the ethical issues that require regulation, Celsi (2023) outlines the difficulty posed by the rapid AI development and the public authorities' ability around the world to keep up with the technology through applicable legislation. The use of deepfake technology, which manipulates videos and audio to create realistic representations of people saying or doing things they did not, presents significant challenges and opportunities from a macromarketing perspective. As a field that studies the interactions between marketing and society, macromarketing needs to consider the ethical, social, and legal implications of this emerging technology. As a meta-theory, Macromarketing analyzes how marketing systems interact with society and reflects on large-scale economic, social, and cultural impacts (Shapiro, 2006 ; Pittz, Steiner & Pennington, 2020 ). In the context of AI and deepfakes, macromarketing offers a critical lens for understanding how these emerging technologies shape consumer behavior, cultural dynamics, and ethical issues. By enabling hyper-realistic representations, deepfakes not only redefine advertising practices but also introduce significant risks. Emerging technologies often generate unforeseen negative externalities, which require robust regulations to protect consumers and social balance (Harris & Carman, 1983 ; Hermann, 2022 ). In the case of recreating public or cultural figures, questions arise about the role of these representations in shaping collective memory and the emotional impact on different audiences. Thus, understanding the implications of using deepfakes through this theoretical perspective is essential to align market practices with ethical and social values. First, the use of deepfakes in marketing can offer new forms of consumer engagement and advertising strategies (Kietzmann, Mills, & Plangger, 2021 ). However, this approach comes with substantial ethical risks, such as the spread of misinformation and the violation of image rights and consent, as already discussed by Harris and Carman ( 1983 ) in the context of marketing regulation. The ability to create realistic deepfakes can make it difficult for consumers to distinguish between reality and manipulation, exacerbating trust and credibility issues in the relationship between brands and consumers (George & George, 2023 ). Furthermore, the use of deepfakes may have broader social and cultural implications. For example, the recreation of public or cultural figures can affect public perception and their legacy, a topic Hutt, Mokwa, and Shapiro ( 1986 ) explore in their discussion of the "parallel political market" in marketing. This matter also raises questions about representation and identity, mainly when deepfakes are used to alter characteristics such as race, gender, or age. Although deepfake technology has impressive applications, its effects on vulnerable populations require special attention. Children, adolescents, and the elderly, for example, have greater difficulty distinguishing between fiction and reality (Turkle, 1997 ). This inability to discern can have psychological impacts, such as the idealization or distortion of digitally recreated public figures, especially when they involve sensitive topics such as grief or nostalgia (Naffi et al., 2023 ). A relevant example is the use of deepfakes for political manipulation, which can exploit vulnerable populations' lack of access to reliable information, thereby increasing the risk of misinformation. This analysis reinforces the need for public policies that account for the impact of these technologies on vulnerable audiences, promoting digital literacy and the development of critical skills. 2.2 Deepfakes regulation The regulation of deepfake technology is still a developing field, with different jurisdictions adopting different approaches to address the ethical and legal challenges it presents. In the United States, for example, initiatives such as the Deepfake Accountability Act (a federal bill that criminalizes the production of deepfakes without identifying that AI generates the content), enacted in 2023, seek to establish a specific legal framework for this technology, requiring the disclosure of manipulated videos or audios and holding creators liable for any damage caused (Chesney & Citron, 2019 ). In the European Union, the approach has been more comprehensive, with the consideration of regulatory frameworks that address various applications of Artificial Intelligence, including deepfakes (European Commission, 2021 ), through the Digital Services Act published in 2022 (European Commission, 2022 ) and the Artificial Intelligence Act (AI Act) signed in early 2024 (European Commission, 2024 ). The European Union's AI Act represents a comprehensive regulatory framework that aims to ensure safety, ethics, and transparency in the AI development and use in member states. By classifying AI systems into risk categories, the European Act prohibits those posing an unacceptable risk and imposes strict obligations on high-risk systems. Although it does not explicitly mention deepfakes, the AI Act addresses manipulative and deceptive techniques that may cover the harmful uses of deepfakes. For high-risk systems and general-purpose AI models, detailed technical documentation, data governance, and transparency measures are required to ensure compliance with ethical and legal guidelines (Future of Life Institute, 2024 ). This evolving regulatory landscape makes cases such as Volkswagen's particularly instructive. In Brazil, where government regulation of deepfakes is still in its infancy, self-regulatory bodies such as CONAR play a crucial role (CONAR, 2020 ). However, between 2010 and 2020, there was a high number of complaints to CONAR related to advertising aimed at children and adolescents. Most of these consumer complaints were dismissed (Matos et al., 2023 ). Six legislative proposals in Brazil address this issue, but as all proposals require approval by both houses of the National Congress, none have been enacted to date. The first initiative was in 2019 and came from the Federal Senate, with Bill No. 5,051/2019, which aims to regulate the AI use in Brazil (Federal Senate, 2024). Also in 2019, the same author of the previous proposal presented Bill No. 5,691/2019, which aimed to establish the National Artificial Intelligence Policy (Federal Senate, 2024). The following year, Bill No. 21/2020 was proposed in the Chamber of Deputies —the so-called Legal Framework for AI —with the intention of establishing principles, rights, duties, and governance instruments for the use of AI in Brazil (Federal Senate, 2024). Even without the approval of any of the previous proposals, in 2021, the Senate received a new proposal, Bill No. 872/2021, whose objective is to provide ethical frameworks and guidelines for the development and use of AI in Brazil (Federal Senate, 2024). In 2023, two new motions were presented, one in the Chamber of Deputies, Bill No. 759/2023, focused on regulating AI systems (Chamber of Deputies, 2024 ), and in the Federal Senate, Bill No. 2,338/2023, focused on AI usage (Federal Senate, 2024). Although providing important guidelines on AI, the Brazilian Bill 2338/2023 contains articles that, lack greater precision and specificity, resulting in subjectivity that could compromise the practical application of the regulations (Chamber of Deputies, 2024 ). The use of vague and adjectival terms leads to inconsistent interpretations, especially of central concepts such as model, impact, and associated risk. These elements require clear, objective definitions for the legislation to be practical and applicable across diverse scenarios. One of the main problems with the text is the absence of objective criteria for classifying an AI system as "high risk." This flaw creates ambiguity and can lead to uneven interpretations. Based on international frameworks, such as the European Union's AI Act, and the discussions of the first OBIA Seminar (Brazilian Observatory of Artificial Intelligence about algorithmic auditing (Meira, 2024 ), the classification of high-risk systems should consider elements such as probability of failure, severity of impact, and scope of automated decisions. If incorporated into the project, these criteria would provide a more solid basis for identifying high-risk systems and targeting specific regulations. Another point of criticism is the lack of attention to the regulation of emerging technologies, such as deepfakes. Despite their obvious risks, the bills under consideration address deepfakes only in the electoral context, as in the Brazilian law project 1.002/2023 (Chamber of Deputies, 2024 ). This approach disregards the dangers of this technology in other areas, including large-scale disinformation, financial fraud, and emotional manipulation. To correct this gap, it would be necessary to include specific regulations addressing deepfakes in commercial and advertising contexts, as well as preventive measures, such as mandatory disclaimers stating that certain content was generated or manipulated by AI and the implementation of tools that guarantee the authenticity of digital content (Nightingale & Wade, 2022 ; Dixon, 2024 ). Dixon, 2024 ). Although the Brazilian Bill 2338/2023 emphasizes fundamental principles such as transparency, security, privacy protection, and respect for ethics and human rights, it fails to detail how these values will be operationalized. For example, while highlighting the promotion of sustainable and inclusive development, it does not present specific measures to protect vulnerable populations from discriminatory or manipulative technologies. One solution would be to detail oversight mechanisms, such as the creation of independent committees, and to establish clear technical standards based on international standards, such as ISO/IEC 23894, which addresses AI governance (ISO/IEC, 2023 ). About governance tools and representative data, Article 18 of Brazilian Bill 2338/2023 requires adequate training, validation, and testing of data, but does not define the objective criteria that such data must meet. Adequate data must be bias-free, up-to-date, and representative of the context in which the system will be applied (Hwang et al., 2021 ). In addition, the automatic recording tools mentioned in the text should include standards for accuracy and robustness, ensuring the traceability and auditability of automated decisions. Overall, Brazilian law project 2338/2023 reflects a legitimate concern with the regulation of AI in Brazil, but it has significant shortcomings. The lack of conceptual clarity, disregard for emerging technologies such as deepfakes, and the guidelines' generality weaken their applicability. Incorporating objective criteria for classifying high-risk systems, expanding the discussion of deepfakes, adopting internationally recognized technical standards, and strengthening oversight mechanisms would be essential steps to improve the legal text and ensure it aligns with society's needs and technological advances. Hence, the issue of deepfake regulation is complex and requires debate and further study, involving interaction between government laws and self-regulatory mechanisms (Whittaker, Letheren & Mulcahy, 2021 ). As technology continues to evolve, a more coordinated and perhaps international approach to addressing ethical challenges is likely to emerge. The impacts of deepfakes undoubtedly warrant further regulatory consideration. However, from the perspectives of literacy and understanding on the part of companies and vulnerable groups, it is essential to intensify research demonstrating how educational techniques and transformative research can be applied to raise consumer awareness of the threats. 2.3 Impact of deepfakes on vulnerable populations The impact of technologies such as deepfakes on vulnerable populations is a critical area that requires attention from regulators, researchers, and educational institutions. Although all segments of the population are subject to the risks of misinformation and digital manipulation, groups such as children, adolescents, the elderly, and individuals with less access to digital education are particularly susceptible. Studies indicate that these populations have greater difficulty distinguishing between real and manipulated content, mainly due to their reliance on unverified sources and lack of digital literacy (Hwang et al., 2021 ; Nightingale & Wade, 2022 ; Naffi et al., 2023 ). The difficulty in recognizing deepfakes not only poses a risk to the spread of misinformation but can also cause significant psychological and social impacts. Children and the elderly, for example, are especially prone to internalizing manipulated information as accurate, which can lead to widespread distrust of legitimate sources of information (Nightingale & Wade, 2022 ). These effects can be amplified in emotionally charged contexts, such as the manipulation of political discourse or the use of deepfakes for fraud and scams. This demonstrates the urgent need for educational and technological measures to mitigate such impacts. One of the most effective strategies for reducing the impact of deepfakes on vulnerable populations is implementing digital literacy programs. Educational programs aimed at children, adolescents, and the elderly, focused on teaching the basic principles of digital manipulation and the characteristics that can expose a deepfake, are essential. Examples of such initiatives are already being tested in educational institutions worldwide (Hwang et al., 2021 ). Although education is crucial, it must be complemented by measures implemented by digital platforms. Companies such as Meta and X (formerly Twitter) need to adopt deepfake-detection technologies to alert users to potential manipulations. In addition, specific regulations requiring clear disclaimers for AI-generated content would be a significant step toward ensuring transparency. In addition to theoretical and technological discussions, there is an urgent need for an integrated approach to mitigate the impacts of deepfakes on vulnerable populations. Education, regulation, and technological solutions must be implemented simultaneously, involving governments, technology companies, educational institutions, and civil society. The role of digital platforms in detecting and flagging manipulated content, combined with digital literacy and blockchain, represents a promising way to protect more vulnerable populations. 3. METHODOLOGY This study adopted the case study approach (Yin, 2014 ), systematically investigating contemporary phenomena within their real-life contexts. Initially, the objectives of the literature review were established to outline the research scope within the context of deepfakes, focusing specifically on public policy and technological innovation. The subsequent phase involved searching databases of recent papers and online materials, such as legislation and media coverage, to ensure the research included relevant studies with diverse perspectives and approaches to the topic. Multiple data sources were used to increase the reliability of the results (Eisenhardt, 1989 ). Documentary research constituted a significant part of the data collection, involving detailed analysis of online materials, including proposed legislation and regulations, academic papers, and media coverage of the application of deepfakes in advertising. This exploration was expanded through a literature review across recognized academic databases, including Google Scholar, EBSCO, and the AI tools Consensus and Typeset. The following keywords were used separately: deepfakes AND "artificial intelligence"; deepfakes AND regulation; and deepfakes AND blockchain for the period from 2016 to March 2024. A total of 32 papers were identified in the databases searched. Figure 1 shows that 2023 had the highest number of publications (10) related to the search terms, likely reflecting the growth of interest in AI, particularly after the launch of OpenAI's ChatGPT in late 2022. The objective of the documentary research and literature review was to ensure a comprehensive investigation of the ethical and legal debates surrounding deepfakes, encompassing contributions from various disciplines to reflect the topic's interdisciplinary nature (Yin, 2014 ). The search and filtering stage consisted of applying the terms established in the databases, followed by selecting papers based on predefined criteria focused on the recent publication period, availability of the complete text, and direct relevance to the impacts of deepfakes. For data extraction, information such as the authors' names, year of publication, study objectives, methodologies used, and main conclusions of each paper were collected. This data collection was instrumental for the subsequent analysis. Source: Prepared by the authors. The AI-based qualitative data analysis tool typeset.io was used to assist in organizing and analyzing the vast amount of material collected, contributing to the integrity and systematization of the literature review results. The critical evaluation of the methodological quality of the selected papers was a crucial step aimed at identifying potential biases and ensuring the reliability of the sources. In addition, efforts were made to mitigate publication bias by examining gray literature and documents in different languages. Through this descriptive and systematic methodology, it was possible to develop a comprehensive picture of the academic debates and political discussions on deepfakes, establishing a robust basis for future research and policy development. 4. DATA ANALYSIS The Data Analysis section is divided into three interrelated subsections that collectively examine the empirical dimension of the research. Subsection 4.1 presents the Volkswagen case, describing the use of deepfake technology to recreate singer Elis Regina in a 2023 advertisement and the ensuing ethical controversy evaluated by CONAR (Brazilian Council for Advertising Self-Regulation). Subsection 4.2 deepens the analysis by examining CONAR’s decision, highlighting tensions between innovation, ethical responsibility, and the protection of vulnerable audiences, while also addressing the absence of clear guidelines on the disclosure of AI-generated content. Finally, subsection 4.3 broadens the discussion by analyzing other recent deepfake cases—including political and corporate incidents—that demonstrate the global diffusion of such technologies and the urgent need for more robust regulatory, educational, and technological mechanisms. Together, these subsections construct an analytical framework that integrates legal, ethical, and social perspectives, elucidating how deepfakes challenge the boundaries between creativity, manipulation, and governance in contemporary digital communication. 4.1 Case description On July 4, 2023, Volkswagen Brazil launched an advertisement using deepfake technology to present a duet between the late Brazilian singer Elis Regina and her daughter, Maria Rita, as shown in Fig. 2 . However, this marketing strategy was met with rejection, leading to its suspension by CONAR due to ethical concerns related to the image use of a deceased person with the family authorization but without the consent of the rights holder (UOL, 2023; G1, 2023; Folha de São Paulo, 2023; Estadão, 2023). Regarding Volkswagen's advertising campaign in Brazil, the issue raised by the complaint (Estadão, 2023) highlights the challenge of distinguishing between fiction and reality in advertising and the possibility that children and adolescents may not readily perceive this distinction. The use of AI to create content that mimics reality with high precision —known as deepfakes —highlights the ability of these representations to influence the perceptions of young audiences. Due to their still-developing cognitive abilities, such audiences may have difficulty discerning the artificial nature of such content, leading to possible misunderstandings or false beliefs. CONAR chose to dismiss the ethical complaint against Volkswagen's advertising campaign, which used deepfake technology to recreate the image of the late singer Elis Regina. This decision was made on August 22, 2023 (Migalhas, 2023 ; Correio Braziliense, 2023; Carta Capital, 2023). Despite the controversy, Volkswagen celebrated the dismissal of the case, and the campaign featuring Elis Regina achieved record views (Livedune, 2023 ). Regarding the issue of image rights, there was unanimous agreement that there was no disrespect to Elis Regina, since her heirs consented to the use of the singer's image, who appeared in the advertisement singing, something she did in life. About image rights, this unanimous decision established that CONAR and Elis Regina's heirs agreed that the recreation was carried out respectfully and ethically. Regarding the impact on vulnerable audiences, especially the need to explicitly inform them about the use of AI in advertisements, the vote was not unanimous. The majority of votes (13 versus 7) found that the AI technique used was evident in the advertising campaign and that no further clarification was needed for the public regarding the use of this technology to recreate Elis Regina's image (Migalhas, 2023 ). Produced by the AlmapBBDO agency, the campaign was considered by Volkswagen to be one of its greatest marketing successes. The campaign not only celebrated an important milestone for the company but also explored an emerging technological innovation, showcasing the singer digitally recreated using deepfake technology, which generated considerable media and public discussion and attention (Automotive Business, 2023 ). As for the campaign results, Volkswagen announced that the commercial achieved a record number of views, highlighting its positive impact on the public. The campaign resonated with the public, divided opinion, and demonstrated the power of technological innovation to create emotionally engaging advertising content (Livedune, 2023 ). Volkswagen's defense of the campaign and disclosure of the figures emphasized how the company valued innovation and creativity, despite the controversies and ethical questions raised by the use of deepfake technology to recreate the image of a deceased figure. 4.2 Analysis of the decision of the advertising council CONAR's (Brazilian Council for Advertising Self-Regulation) decision on Volkswagen's advertising campaign reflects several complexities inherent in emerging advertising and technology practices. In the ethical and regulatory sphere, the decision first addresses image rights. According to Faden, Beauchamp, and Kass ( 2014 ), image rights are a fundamental ethical principle, particularly in cases of digital recreation of deceased individuals. CONAR's decision, which found no violation of Elis Regina's image rights with her heirs' consent, is aligned with the guidelines for informed consent and respect for image rights (Lessig, 1999). However, analysis of the decision reveals significant gaps in the treatment of vulnerable audiences and the need for greater transparency. Evidence suggests that vulnerable groups, such as children and the elderly, may have difficulty distinguishing reality from the digital manipulation presented in the commercial, especially given the sophistication of deepfake technologies (Turkle, 1997 ; Scherer & Pennycook, 2020 ). This cognitive limitation, combined with the emotional impact of seeing public figures recreated so realistically, raises concerns about possible psychological and social effects. Studies indicate that the confusion generated by hyper-realistic digital content can distort perceptions about the legacy of the recreated person, exacerbating the risk of unrealistic idealizations or even trivializing their memory (Naffi et al., 2023 ). Furthermore, the debate at CONAR on the impact of advertising on vulnerable audiences raises the question of the duty to explicitly inform about the use of AI in the production of the ad. The decision that the technique's use was evident was not unanimous, highlighting the vulnerability of certain groups to misinformation —an issue that intensifies with age and trust in unverified sources (Scherer & Pennycook, 2020 ). In this context, CONAR's decision not to explicitly require a disclaimer stating that the campaign used AI and deepfake techniques reveals a missed opportunity to set a more robust ethical precedent. The use of clear and visible disclaimers could not only educate the public about the technological nature of the content but also protect vulnerable groups from confusion and promote greater transparency in advertising practices. This preventive measure is widely advocated in the literature as an essential approach to mitigating the risks associated with misinformation and digital manipulation (Mirsky & Lee, 2020). Adopting mandatory guidelines to include disclaimers in campaigns that use advanced technologies, such as AI, could better align advertising practices with the principles of self-regulation and ethical responsibility. Furthermore, such an approach would align with international efforts to regulate emerging technologies, such as those proposed in the European Union's AI Act, which emphasize transparency and consumer protection in vulnerable situations. CONAR's decision highlights the importance of moving forward with building a more comprehensive regulatory framework that considers not only the legal and ethical aspects but also the social and cultural impacts of technologies such as AI. The inclusion of measures such as the mandatory use of disclaimers could serve as a model for proactive regulation, promoting more responsible advertising and adequately protecting consumers in an increasingly complex digital environment. In this context, CONAR's decision underscores the importance of advertising self-regulatory bodies in a regulatory landscape where government regulations are still under development. However, the decision suggests the need for a more holistic regulatory framework that incorporates the social and cultural impacts of technologies such as artificial intelligence, going beyond mere legal and ethical considerations (Post, 2000). It is essential to note that this overview of CONAR's decision reflects on how organizations should navigate the use of emerging technologies, balancing innovation and creativity with ethical and legal responsibility. The ease with which this technology can be used for unethical and malicious purposes, such as spreading misinformation and falsifying identities, has been a growing concern (Mirsky & Lee, 2020; Grewal, Guha, & Becker, 2024 ). 4.3 Analysis of other deepfake cases and recent events Social media has often been the scene of cases that illustrate the urgency of more robust regulation. Notable examples include the fake images of the Pope wearing a stylish white coat and the fake arrest of Donald Trump (Metropoles, 2023 ), both in 2023. In early 2024, an employee of a multinational company transferred $ 25 million to criminals after a video call with a Chief Financial Officer produced by a deepfake (CNN, 2024). These incidents highlight the need for a more comprehensive regulatory approach to mitigate the risks posed by irresponsible use of technologies such as deepfakes. The proliferation of AI-generated images depicting fake events raises significant concerns about people's ability to discern their authenticity. Not everyone has the technical knowledge or attention to detail necessary to identify the subtleties that distinguish a real image from a fake one. This poses a serious risk, as misleading images can spread misinformation, influence public opinion, and even manipulate political and social events. The difficulty in verifying image authenticity can lead to the spread of false information, undermining trust in digital content and negatively impacting society (Kietzmann et al., 2020 ). In February 2024, OpenAI announced its new AI called Sora. OpenAI's Sora can generate realistic, creative videos from textual instructions. With the ability to create videos up to one minute long, Sora maintains visual fidelity and precisely follows user requests, generating complex scenes that include multiple characters, specific movements, and accurate details of scenery and objects (OpenAI, 2024 ). As of March 2024, the platform was not available to the public because, according to OpenAI ( 2024 ), the company was conducting tests and developing tools to detect misleading content, aiming to prevent misinformation, hateful content, and bias. However, Sora does not fall into the category of explainable AI (XAI), which raises questions about its functioning and transparency, as well as the potential risk of generating deepfakes on the platform. XAI refers to an interpretable system that provides a comprehensible explanation of its output and draws attention to the ethics of AI, focusing on how AI systems make decisions, the explanations for those decisions, and how those decisions are communicated to relevant stakeholders (Vainio-Pekka et al., 2023 ). The advancement of deepfakes has brought serious challenges, as evidenced in the 2024 elections in Brazil. The Superior Electoral Court (TSE) faced significant problems with digitally manipulated videos, which harmed candidates by spreading false information. In response, the TSE regulated the use of AI in election campaigns, expressly prohibiting the creation and propagation of deepfakes. In addition, it determined that the AI use in campaigns should be accompanied by an explicit warning to promote greater transparency (Superior Electoral Court, 2024 ). One of the most notorious cases involved São Paulo mayoral candidate Tabata Amaral, who was targeted by fake videos. Digitally manipulated images associating her with sexual content were posted on social media to damage her public image. In response, the candidate took legal action through the Electoral Court to hold those involved accountable and combat misinformation during the election period (UOL Notícias, 2024). Starlink also faced challenges related to the use of deepfakes. Scammers created fraudulent ads promoting fake lifetime internet plans, using videos that cloned the voice of Brazilian congressman Nikolas Ferreira. This misleading content tricked consumers into visiting fraudulent websites, compromising the security of their personal and financial data and damaging the company's reputation (Boatos.org, 2024 ). These cases highlight the risks of malicious AI use and reinforce the importance of regulatory, legal, and technological measures to combat deepfakes, protecting the integrity of electoral processes and the reputations of individuals and organizations. In this sense, the use of Blockchain may be an appropriate alternative. 5. DISCUSSION Considering the technological nature of this paper, the discussion section emphasizes the proposition of a solution to the identified issues. Hence, blockchain emerges as a promising technology to mitigate the problems associated with deepfakes. Through its decentralized, immutable structure, blockchain can authenticate and track the origin of digital content, increasing confidence in the material disclosed. Dixon ( 2024 ) points out that blockchain's transparent record is essential to ensuring media integrity, while Seneviratne ( 2022 ) suggests that its integration with AI can offer robust solutions for verifying content authenticity. In the context of advertising and the ethical use of AI, blockchain could ensure that digital recreations, as in the Volkswagen case, are appropriately labeled and verified for origin, reducing the risk of confusion between reality and fiction. This technological solution not only addresses ethical issues but also strengthens consumer confidence in AI-based marketing practices. Known for its immutable, transparent transaction record, blockchain could be decisive in authenticating the origin and maintaining the integrity of multimedia content. Dixon ( 2024 ) suggests that blockchains, by providing an unalterable audit trail, help identify the source of digital content —a crucial element in combating the spread of fake videos. Seneviratne ( 2022 ) also argues that blockchain technology can significantly reduce the prevalence of deepfakes by ensuring the authenticity and integrity of visual media. By leveraging the decentralized nature, encryption capabilities, and tamper-proof features of blockchain, a secure and reliable system can be established to verify the authenticity of content. The author argues that by combining blockchain technology with AI techniques —such as tamper-proof blockchain and AI models to detect fake news —a synergistic approach can be developed to combat the proliferation of misleading information and deepfakes across various platforms. Blockchain-based technologies offer a promising approach to verifying the authenticity of digital media, playing a crucial role in combating the proliferation of deepfakes. Chan et al. ( 2020 ) highlight blockchain's ability to provide a decentralized architecture that enables the verification of digital content authenticity and the tracking of its historical provenance. This feature is especially pertinent, as deepfakes are becoming increasingly sophisticated and challenging to identify. Additionally, Prajapati et al. ( 2023 ) suggest using blockchain to authenticate products through linked QR codes, reinforcing trust and authenticity and, consequently, mitigating the circulation of counterfeit products. This approach demonstrates the versatility of blockchain beyond cryptocurrencies, expanding its use to digital security across the broad spectrum. Jing and Murugesan ( 2021 ) complement this view, proposing that blockchain can effectively reduce the incidence of deepfakes by implementing a trust index model on social media. This model prioritizes data privacy, credibility, and transparency, creating a more secure and reliable digital environment where false information and media manipulation are readily identifiable and discredited. In the context of deepfakes, a key aspect is the importance of double-checking information upon receipt, especially given the significant challenge of accurately detecting them. The ability to verify information from a reliable source is crucial, especially in educational processes that aim to protect vulnerable groups, such as children and the elderly. Deepfakes have great potential for negative impact, given their ability to imitate reality compellingly. Automated detection of deepfakes is essential to prevent the spread of false information that can manipulate public opinion, compromise electoral processes, and cause profound social and economic damage (Verdoliva, 2020 ). In addition, Shahzad et al. ( 2022 ) emphasize the importance of combining technological solutions with digital education and awareness to mitigate the harmful effects of this technology. Another relevant point in tackling deepfakes is the use of blockchain as a supporting technology. Blockchain offers several improvements that should be explored, especially for certifying the creator's identity and the authenticity of content. This technology can provide greater security through immutable and decentralized records, creating a reliable validation system for digital publications (Sudhakar & Shanthi, 2023 ). Heidari et al. ( 2023 ) argue that NFTs (Non-Fungible Tokens), associated with blockchain, can be an effective tool for authenticating content, ensuring origin and credibility through immutable records. Sam Altman, from OpenAI, proposed the WorldID platform, which uses blockchain to ensure individuals' universal identity. This proposal has the potential to reduce the spread of false information by eliminating anonymity, as people who spread fake news would have no way to hide (Singh & Dhiman, 2023). In addition, decentralized governance systems can be integrated with technologies such as oracles, allowing users to attest to the authenticity of information, thereby promoting a more resilient ecosystem against misinformation (Gupta et al., 2023 ). This integration of blockchain systems with external platforms facilitates the creation of reputation rankings for content and its authors, validating reliability through distributed governance mechanisms. As noted by Singh & Dhiman (2023), decentralization mitigates the harms of a centralized authority with complete control over reputation, thereby enhancing system resilience and reliability. Hasan et al. ( 2024 ) proposes an innovative solution based on blockchain and NFTs to address the challenges posed by deepfake technology, particularly in the metaverse and digital media. The solution uses NFTs to authenticate and protect the originality of digital content. Digital content creators register on an NFT marketplace (NFTM), where they can create and trade digital assets linked to NFTs. The blockchain ensures the immutability of records and the provenance of data, while smart contracts enable secure, traceable transactions. In addition, the proposal integrates oracles to update dynamic NFT metadata, allowing the status of digital assets to be adjusted based on external data (Fig. 3 ). To mitigate the high costs of on-chain storage, NFT metadata is stored in a decentralized manner using systems such as IPFS (Interplanetary File System), with only the file hash stored on the blockchain. This model reinforces data security and integrity while ensuring the authenticity and reliability of content through a decentralized reputation system that assigns "trust badges" to creators based on their trustworthiness. With this approach, the paper proposes a robust method for addressing the risks associated with deepfakes, ensuring the authenticity, ownership, and originality of digital assets. A multifaceted approach combining regulation, technology, and educational processes is needed to effectively combat the challenges posed by deepfakes. Shoaib et al. ( 2023 ) emphasize that public policies aligned with technological solutions, such as blockchain integration, are essential to mitigate social, political, and economic impacts. Complementarily, Rana et al. ( 2022 ) emphasize that digital education and the development of automated detection tools are indispensable for reducing the impact of these threats in the contemporary world. Source: Adapted from Hasan et al. ( 2024 ) The triangulation of these perspectives highlights the applicability of blockchain as a robust tool against misinformation. The synergy between blockchain decentralization, content and product authentication, and the creation of a trust index on social media establishes a theoretical and practical framework for mitigating the challenges posed by deepfakes. Therefore, adopting this emerging technology can help ensure information integrity in the digital age, paving the way for a technological response to the advancement of media manipulation techniques. 6. CONCLUSION Although efforts are being made to establish ethical and legal guidelines, the Volkswagen case and CONAR's decision indicate that the issue should be discussed further, given the incipient nature of laws and regulations addressing the use of emerging technologies, such as AI in advertising and deepfakes. Discussions on regulation in 2024 should be deepened to address not only legal issues but also the social and cultural impacts of these technologies (Post, 2000; Kietzmann, Mills, & Plangger, 2020; Celsi, 2023). The use of emerging technologies, such as AI and deepfakes, represents both an opportunity for innovation and a significant ethical and regulatory challenge. This technological paper sought to investigate the ethical, social, and cultural implications of deepfakes in advertising. The analysis highlighted the gap between technological advancement and regulation that could mitigate the negative impacts of these practices, offering reflections on consent, protection of vulnerable audiences, cultural impacts, and practical solutions, such as blockchain. In addition, the paper contributes by addressing the Brazilian and international regulatory landscape, presenting a compilation of the main laws under discussion and the bills currently being debated in the Brazilian Congress. This analysis included a detailed review of existing and proposed regulations governing the use of deepfakes and other emerging technologies across commercial, cultural, and social contexts. There is a noticeable gap between rapid technological advancement and the ability of laws to keep pace with these changes. In Brazil, among the bills mentioned, law project 2630/2020, known as the "Fake News Law," stands out. It aims to establish guidelines for the transparency of digital platforms and accountability for false or manipulated content. Another relevant bill is law project 2338/2023, which proposes a regulatory framework for the protection of personal data and privacy, including the use of AI technologies. The data presented show that current regulatory decisions, such as those made by CONAR, still struggle to keep pace with technological development. Although the consent of Elis Regina's heirs ensured respect for image rights, the absence of clear guidelines on transparency, such as the requirement for disclaimers informing the use of deepfakes, reveals a missed opportunity to protect consumers, especially the most vulnerable. Recent cases, such as the deepfakes disseminated during the 2024 elections and scams involving voice cloning to promote fake Starlink plans, reinforce the need for more agile and effective regulations. In addition, the examples analyzed demonstrate that hyper-realistic digital manipulation can cause confusion between fiction and reality, especially among children, adolescents, and the elderly, who have greater difficulty discerning the nature of the content. This confusion can have emotional and cultural impacts, such as distorting public perception of digitally recreated historical or cultural figures. Therefore, the mandatory inclusion of disclaimers for the use of technologies such as blockchain to authenticate digital content presents a practical and ethical solution to mitigate these risks, strengthening public confidence and promoting more responsible business practices. This study has some limitations, including its focus on a single case and its lack of a comprehensive analysis of diverse legal and cultural contexts. Future contributions could include exploring the deepfake phenomenon from other empirical studies, as well as a more in-depth analysis of how legislation should address ethical and social issues related to the use of deepfakes, especially when they involve deceased persons (Chesney & Citron, 2019 ; Rini, 2020 ; Kietzmann, Mills, & Plangger, 2020). It would be relevant to investigate how advertising could incorporate clear messages about the use of deepfakes. Issues regarding the protection of vulnerable populations are also crucial (Naffi et al., 2023 ). Mechanisms could be developed to help identify deepfakes and promote open debate in society about their impact. In addition, using blockchain to verify the authenticity of videos from reputable sources, such as established media outlets or recognized institutions, would introduce a valuable tool for distinguishing between factual and fabricated content. Dixon ( 2024 ) argues that blockchain can add a layer of security and authentication to digital content, helping restore trust in information shared online. Linking technologies such as blockchain to mandatory transparency practices, such as the use of explicit disclaimers informing the use of deepfakes, reinforces both the credibility of the content and consumer protection. This approach allows companies to innovate responsibly, promoting confidence in the use of advanced technologies. Future research could explore how deepfakes impact sectors beyond advertising, such as journalism and entertainment, and how different regulatory approaches can mitigate their risks. In addition, empirical studies could investigate public perception of deepfakes and the effectiveness of solutions such as disclaimers and blockchain in protecting against digital manipulation. Another aspect that deserves investigation is the role of marketing managers' ethical awareness in the use of these technologies. What are the accepted ethical limits in the market? How can companies incorporate governance practices that prioritize transparency and consumer protection? These issues are fundamental to building a more responsible market that is aligned with social values. Finally, bills on the subject could benefit from a broader discussion involving not only Congress but also civil society to promote greater awareness of the impacts of this technology. The use of deepfakes and AI in marketing poses significant challenges that require a collective effort among companies, regulatory agencies, governments, and other actors. Self-regulatory bodies, such as CONAR, play an important role in maintaining ethical standards, but the need for more comprehensive and up-to-date government laws and regulations is undeniable. The speed of technological innovation requires an equally agile regulatory response that considers the multiple impacts of these technologies on different segments of society. The future of these AI-based technologies will depend, in large part, on the ability to regulate and use them in ways that respect individual rights and promote collective benefits. In short, the future of digital manipulation technologies depends on our ability to balance innovation with ethical responsibility, protecting the rights of all citizens, especially the most vulnerable. Declarations Author Contribution D.D.G. wrote the main manuscript text, R.B.C. reviewed and wrote some paragraphs of background, introduction, and discussion, and R.S.L. and H.C.N. reviewed the manuscript. References Automotive Business. (2023). Volkswagen defends campaign with Elis Regina. Automotive Business . https://www.automotivebusiness.com.br BBC. (2023, March 23). Fake Trump arrest photos: How to spot an AI-generated image. BBC News . https://www.bbc.com/news/world-us-canada-65069316 Boatos.org. (2024). A scam uses Nikolas Ferreira's voice, dubbed with AI, to talk about Starlink. Boatos.org . https://www.boatos.org/tecnologia/nikolas-ferreira-anunciou-que-elon-musk-vai-dar-starlink-vitalicia-por-r-184.html Chamber of Deputies. (2023). The proposal regulates the use of artificial intelligence . https://www.camara.leg.br/noticias/968967-proposta-regulamenta-utilizacao-da-inteligencia-artificial Chamber of Deputies. (2024). Bill No. 1,002/2023 . https://www.camara.leg.br/propostas-legislativas/2350539 Chamber of Deputies. (2024). Bill No. 759/2023 . https://www.camara.leg.br/propostas-legislativas/2349685 Chan, C. C. K., Kumar, V., Delaney, S., & Gochoo, M. (2020). Combating deepfakes: Multi-LSTM and blockchain as proof of authenticity for digital media. In 2020 IEEE/ITU International Conference on Artificial Intelligence for Good (AI4G) (pp. 55–62). IEEE. https://doi.org/10.1109/AI4G50087.2020.9319390 Chesney, R., & Citron, D. (2019). Deep fakes: A looming challenge for privacy, democracy, and national security. California Law Review , 107(6), 1753–1819. https://doi.org/10.15779/Z38R49J16H CNN. (2024, February 4). Finance worker pays out $25 million after video call with deepfake “chief financial officer”. CNN International . https://edition.cnn.com/2024/02/04/asia/deepfake-cfo-scam-hong-kong-intl-hnk/index.html CONAR. (2020). Brazilian Code of Advertising Self-Regulation . CONAR. DataSciConnect. (2023). Generative AI deepfakes: Volkswagen ad revives Brazilian singer, raises concerns. DataSciConnect . https://datasciconnect.com/generative-ai-deepfakes-volkswagen-ad-revives-brazilian-singer-raises Dixon, C. (2024). Read and write your own . Random House. Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review , 14(4), 532–550. https://doi.org/10.2307/258557 European Commission. (2021). White paper on artificial intelligence: A European approach to excellence and trust . Publications Office of the European Union. European Commission. (2022). The Digital Services Act . https://commission.europa.eu/strategy-and-policy/priorities-2019-2024/europe-fit-digital-age/digital-services-act_en European Commission. (2024). EU Artificial Intelligence Act . https://artificialintelligenceact.eu Faden, R. R., Beauchamp, T. L., & Kass, N. E. (2014). A history and theory of informed consent . Oxford University Press. https://doi.org/10.1093/med/9780190495679.001.0001 Folha de S.Paulo. (2023). Deepfake failed in a commercial with Elis Regina by ignoring her past. Folha de S.Paulo . https://www1.folha.uol.com.br/colunas/manuela-cantuaria/2023/07/a-tecnologia-deepfake-falhou-no-comercial-com-elis-regina.shtml Furze, L. (2024a, August 23). AI images are already more realistic than you think. Leon Furze Blog . https://leonfurze.com/2024/08/23/ai-images-are-already-more-realistic-than-you-think/ Furze, L. (2024b). Deepfake game. Leon Furze Blog . https://leonfurze.com/deepfake-game/ Future of Life Institute. (2024). The act texts | EU Artificial Intelligence Act . https://artificialintelligenceact.eu/the-act/ George, A. S., & George, A. S. H. (2023). Deepfakes: The evolution of hyper-realistic media manipulation. Partners Universal Innovative Research Publication , 1(2), 48–54. https://doi.org/10.5281/zenodo.10148558 Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., iCourville, A., & Bengio, Y. (2014). Generative adversarial nets.In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, & K. Q. Weinberger (Eds.), Advances in neural information processing system s (Vol. 27) . http://arxiv.org/abs/1406.2661 Grewal, D., Guha, A., & Becker, M. (2024). AI is changing the world: For better or for worse? Journal of Macromarketing , 44(4), 870–882. https://doi.org/10.1177/02761467241254450 Gupta, G., Raja, K., Gupta, M., Jan, T., Whiteside, S. T., & Prasad, M. (2023). A comprehensive review of deepfake detection using advanced machine learning and fusion methods. Electronics , 12(4), 865. https://doi.org/10.3390/electronics12040865 ISO/IEC. (2023). ISO/IEC 23894:2023 – Information technology — Artificial intelligence —Guidance on risk management . International Organization for Standardization. Ham, C.-D. (2017). Exploring how consumers cope with online behavioral advertising. International Journal of Advertising , 36(4), 632–658. https://doi.org/10.1080/02650487.2016.1239878 Harris, R. G., & Carman, J. M. (1983). Public regulation of marketing activity: Part I: Institutional typologies of market failure. Journal of Macromarketing , 3(1), 49–58. Hasan, H. R., Salah, K., Jayaraman, R., Yaqoob, I., & Omar, M. (2024). NFTs to combat deepfakes and fake metaverse content. Internet of Things , 25, 101133. https://doi.org/10.1016/j.iot.2024.101133 Heidari, A., Navimipour, N. J., Dag, H., & Unal, M. (2023). Deepfake detection using deep learning methods: A systematic and comprehensive review. WIREs Data Mining and Knowledge Discovery , 13(6), e1528. https://doi.org/10.1002/widm.1528 Hermann, E. (2022). Leveraging artificial intelligence in marketing for social good—An ethical perspective. Journal of Business Ethics , 179(1), 43–61. https://doi.org/10.1007/s10551-021-04843-y Hutt, M. D., Mokwa, M. P., & Shapiro, S. J. (1986). The politics of marketing: Analyzing the parallel political marketplace. Journal of Marketing , 50(1), 40–51. Hwang, Y., Kim, H., & Rho, J. (2021). Deepfakes and scientific knowledge dissemination. Scientific Reports , 11, 17769. https://doi.org/10.1038/s41598-021-96343-0 InternetLab. (2017). Black Mirror: “Be Right Back” (S02 E01), mourning and digital inheritance. InternetLab . https://internetlab.org.br/pt/opiniao/black-mirror-be-right-back-s02-e01-luto-e-heranca-digital/ Jing, T. W., & Murugesan, R. K. (2021). Protecting data privacy and preventing fake news and deepfakes in social media via blockchain technology. In Advances in Cyber Security: ACeS 2020 (pp. 674–684). Springer. https://doi.org/10.1007/978-981-15-9129-7_48 Kietzmann, J., Lee, L., McCarthy, I., & Kietzmann, T. (2020). Deepfakes: Trick or treat? Business Horizons , 63(2), 135–146. https://doi.org/10.1016/j.bushor.2019.11.006 Kietzmann, J., Mills, A., & Plangger, K. (2021). Deepfake: Perspectives on the future “reality” of advertising and branding. International Journal of Advertising , 40(3), 473–485. https://doi.org/10.1080/02650487.2020.1834211 Kshetri, N., Dwivedi, Y. K., Davenport, T. H., & Panteli, N. (2024). Generative artificial intelligence in marketing: Applications, opportunities, challenges, and research agenda. International Journal of Information Management , 75, 102716. https://doi.org/10.1016/j.ijinfomgt.2023.102716 Kumar, V., Kotler, P., Gupta, S., & Rajan, B. (2024). EXPRESS: Generative AI in marketing: Promises, perils, and public policy implications. Journal of Public Policy & Marketing . https://doi.org/10.1177/07439156241286499 Lacerda, I., & Mattos, T. R. (2020). Be Right Back: Humans, artificial intelligence, and Dasein in Black Mirror. Communications, technologies et développement . https://doi.org/10.4000/ctd.3353 Lessig, L. (2006). Code: And other laws of cyberspace (Version 2.0) . Basic Books. Livedune. (2023). 70 years of Volkswagen in Brazil: Campaign features Elis Regina. Livedune Blog . https://livedune.com/br/blog/70-anos-de-volkswagen-no-brasil-campanha-elis-regina/ Matos, J. de P., Gondo, M., Mota, L. S. E., & Horta, P. M. (2023). Food advertising aimed at children and adolescents in Brazil: Longitudinal analysis of complaints to CONAR. Ciência & Saúde Coletiva , 28(7), 1959–1970. https://doi.org/10.1590/1413-81232023287.14752022 Mazis, M. B. (1997). Marketing and public policy: Prospects for the future. Journal of Public Policy & Marketing , 16(1), 139–143. https://doi.org/10.1177/074391569701600112 Mazis, M. B., Settle, R. B., & Leslie, D. C. (1977). Elimination of phosphate detergents and psychological reactance. Journal of Marketing Research , 14(3), 390–395. https://doi.org/10.1177/002224377701400315 Meira, W. (2024). Lecture on Artificial Intelligence and Governance [Video]. YouTube. https://www.youtube.com/live/O9czIDgeVFI Metropoles. (2023). Fake images of the Pope wearing a stylish coat go viral on social media. Metrópoles . https://www.metropoles.com/mundo/religiao-int/imagens-falsas-do-papa-com-casaco-estiloso-viralizam-nas-redes-sociais Migalhas. (2023). Conar files lawsuit against Volks for commercial featuring Elis Regina. Migalhas . https://www.migalhas.com.br/quentes/392385/conar-arquiva-processo-contra-volks-por-comercial-com-elis-regina Mittelstadt, B. D., Allo, P., Taddeo, M., Wachter, S., & Floridi, L. (2016). The ethics of algorithms: Mapping the debate. Big Data & Society , 3(2), 1–21. https://doi.org/10.1177/2053951716679679 Mirsky, Y., & Lee, W. (2021). The creation and detection of deepfakes: A survey. ACM Computing Surveys , 54(1), 1–41. https://doi.org/10.1145/3392209 Motta, G. da S. (2022). What is a technological paper? Revista de Administração Contemporânea , 26(Sup. 1), e220208. https://doi.org/10.1590/1982-7849rac2022220208.por Naffi, N., Charest, M., Danis, S., Pique, L., Davidson, A., Brault, N., Bernard, M., & Barma, S. (2023). Empowering youth to combat malicious deepfakes and disinformation: An experiential and reflective learning experience informed by personal construct theory. Journal of Constructivist Psychology , 1–22. https://doi.org/10.1080/10720537.2023.2294314 Nightingale, S. J., & Wade, K. A. (2022). Identifying and minimizing the impact of fake visual media: Current and future directions. Memory, Mind & Media , 1, e15. https://doi.org/10.1017/mem.2021.15 OpenAI. (2024). Sora . https://openai.com/sora Pittz, T. G., Steiner, S. D., & Pennington, J. R. (2020). An ethical marketing approach to wicked problems: Macromarketing for the common good. Journal of Business Ethics , 164, 301–310. https://doi.org/10.1007/s10551-019-04277-7 Prajapati, S., Gadhari, J., Sawant, T., Kini, J., & Solanki, S. (2023). Strengthening supply chain integrity with blockchain-based anti-counterfeiting measures. In 2023 International Conference on Innovative Data Communication Technologies and Application (ICIDCA) (pp. 786–790). IEEE. Rana, M., Nobi, M. N., Murali, B., & Sung, A. H. (2022). Deepfake detection: A systematic literature review. IEEE Access , 10, 25494–25513. https://doi.org/10.1109/ACCESS.2022.3154404 Ricciardi Celsi, L. (2023). The dilemma of rapid AI advancements: Striking a balance between innovation and regulation by pursuing risk-aware value creation. Information, 14 (12), 645. https://doi.org/10.3390/info14120645 Rini, R. (2020). Deepfakes and the epistemic backstop [Preprint]. PhilArchive. https://philarchive.org/rec/RINDAT-2 Shahzad, H. F., Rustam, F., Flores, E. S., Vidal Mazón, J. L., de la Torre Díez, I., & Ashraf, I. (2022). A review of image processing techniques for deepfakes. Sensors, 22 (12), 4556. https://doi.org/10.3390/s22124556 Scherer, L. D., & Pennycook, G. (2020). Who is susceptible to online health misinformation? American Journal of Public Health , 110(S3), S276–S277. https://doi.org/10.2105/AJPH.2020.305908 Seneviratne, O. (2022). Blockchain for social good: Combating misinformation on the web with AI and blockchain. In Proceedings of the 14th ACM Web Science Conference 2022 (pp. 435–442). ACM. https://doi.org/10.1145/3501247.3531549 Shapiro, S. J. (2006). Macromarketing: Origins, development, current status, and possible future direction. European Business Review , 18(4), 307–321. https://doi.org/10.1108/09555340610677449 Shoot Online. (2023, August 15). The best work you may never see: Volkswagen and AlmapBBDO reunite the mother. Shoot Online . https://www.shootonline.com/video/best-work-you-may-never-see-volkswagen-almapbbdo-reunite-mother-and-daughter-duet-ai Shoaib, M. R., Wang, Z., Ahvanooey, M. T., & Zhao, J. (2023). Deepfakes, misinformation, and disinformation in the era of frontier AI, generative AI, and large AI models. In 2023 International Conference on Computer and Applications (ICCA) (pp. 1–7). IEEE. https://doi.org/10.1109/ICCA58330.2023.10342321 Sudhakar, K. N., & Shanthi, M. B. (2023). Deepfake: An endanger to cyber security. In 2023 International Conference on Sustainable Computing and Smart Systems (ICSCSS) (pp. 1542–1548). IEEE. https://doi.org/10.1109/ICSCSS56776.2023.10170989 Superior Electoral Court. (2024). TSE prohibits the use of artificial intelligence to create and spread false content in elections . https://www.tse.jus.br/comunicacao/noticias/2024/Fevereiro/tse-proibe-uso-de-inteligencia-artificial-para-criar-e-propagar-conteudos-falsos-nas-eleicoes The Drum. (2023, August 15). Heartwarming tribute or ethical breach? AI-powered Volkswagen ad sparks controversy. The Drum . https://www.thedrum.com/news/2023/08/15/heartwarming-tribute-or-ethical-breach-ai-powered-volkswagen-ad-sparks-controversy Turkle, S. (1997). Life on the screen: Identity in the age of the Internet . Simon & Schuster. UOL. (2023, July 14). Elis, rights of the dead, and AI: Will Conar set limits for advertising? UOL Economia — Mídia e Marketing . https://economia.uol.com.br/newsletters/midia-e-marketing/2023/07/14/julgamento-de-comercial-com-elis-pode-ser-historico-para-a-publicidade.htm UOL News. (2024, September 14). Tabata Amaral takes legal action after photos are manipulated and published on social media. UOL Notícias . https://noticias.uol.com.br/ultimas-noticias/agencia-estado/2024/09/14/tabata-amaral-aciona-a-justica-por-crime-de-injuria-eleitoral.htm Vainio-Pekka, H., Agbese, M., Jantunen, M., Vakkuri, V., Mikkonen, T., Rousi, R., & Abrahamsson, P. (2023). The role of explainable AI in the research field of AI ethics. ACM Transactions on Interactive Intelligent Systems , 13(4), Article 67. https://doi.org/10.1145/3599974 Verdoliva, L. (2020). Media forensics and deepfakes: An overview. IEEE Journal of Selected Topics in Signal Processing , 14(5), 910–932. https://doi.org/10.1109/JSTSP.2020.2972392 Whittaker, L., Letheren, K., & Mulcahy, R. (2021). The rise of deepfakes: A conceptual framework and research agenda for marketing. Australasian Marketing Journal , 29(3), 204–214. https://doi.org/10.1177/1839334921999479 Yin, R. K. (2014). Case study research: Design and methods (5th ed.). Sage Publications. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8118450","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":551222550,"identity":"6c33c068-865b-4b4a-a1cc-3c7e9758fce1","order_by":0,"name":"Diocélio Dornela Goulart","email":"data:image/png;base64,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","orcid":"","institution":"Fundação Dom Cabral","correspondingAuthor":true,"prefix":"","firstName":"Diocélio","middleName":"Dornela","lastName":"Goulart","suffix":""},{"id":551222551,"identity":"4174517a-fe0a-4074-987b-0c21ba5d4174","order_by":1,"name":"Rodrigo Baroni de Carvalho","email":"","orcid":"","institution":"Pontifícia Universidade Católica de Minas Gerais","correspondingAuthor":false,"prefix":"","firstName":"Rodrigo","middleName":"Baroni","lastName":"de Carvalho","suffix":""},{"id":551222552,"identity":"a6334b33-81d2-4a89-b270-54349ac1d3f1","order_by":2,"name":"Ramon Silva Leite","email":"","orcid":"","institution":"Pontifícia Universidade Católica de Minas Gerais","correspondingAuthor":false,"prefix":"","firstName":"Ramon","middleName":"Silva","lastName":"Leite","suffix":""},{"id":551222553,"identity":"df1be3d2-c6b5-4db1-9cad-b85efe4d94dc","order_by":3,"name":"Henrique de Castro Neves","email":"","orcid":"","institution":"Fundação Dom Cabral","correspondingAuthor":false,"prefix":"","firstName":"Henrique","middleName":"de Castro","lastName":"Neves","suffix":""}],"badges":[],"createdAt":"2025-11-14 23:23:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8118450/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8118450/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97668689,"identity":"e5c4151e-81e6-43e5-89a5-d8dda5adf872","added_by":"auto","created_at":"2025-12-08 09:26:04","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1444122,"visible":true,"origin":"","legend":"","description":"","filename":"2025deepfakesimplicationsreviewed.docx","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/2853adcfc6c20d10993a45c6.docx"},{"id":97460199,"identity":"5f2f3fbc-add7-471c-b6d8-756a80d0ded8","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5601,"visible":true,"origin":"","legend":"","description":"","filename":"ebb5f2c876164c8f8409c4504e097eaa.json","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/7e2428d67ee7984f5e9641e5.json"},{"id":97460202,"identity":"2568f53a-e22e-4526-b74d-e24575634529","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":154391,"visible":true,"origin":"","legend":"","description":"","filename":"ebb5f2c876164c8f8409c4504e097eaa1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/f937b3025696c53ac40a6618.xml"},{"id":97668411,"identity":"cd38a0fe-7965-43ec-85eb-ad8d78a4b6b7","added_by":"auto","created_at":"2025-12-08 09:25:30","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":52532,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/24a54ccd5c494d3e17eaf87d.png"},{"id":97460200,"identity":"e05a947f-2183-4418-91fc-f753ece46b51","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":59004,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/e5adb468209b1ef21deac885.png"},{"id":97460207,"identity":"be84efc8-fb6e-4e0f-a932-f638c64c61b7","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40921,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/e6ccdc77c1246ed2457bbd13.png"},{"id":97460208,"identity":"b35576d3-0fc9-4585-a0af-490fa6b29f5c","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"xml","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":153145,"visible":true,"origin":"","legend":"","description":"","filename":"ebb5f2c876164c8f8409c4504e097eaa1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/5f1bf3a48c63fbe583c01db2.xml"},{"id":97460210,"identity":"44079ca4-737c-4177-aaf8-b8ca10735ff6","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"html","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":163538,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/643eb8ae07c60bc6c2e9ae01.html"},{"id":97668522,"identity":"f1a941ad-94de-4aa3-bb0e-83b69b3da48e","added_by":"auto","created_at":"2025-12-08 09:25:43","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":55613,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eNumber of papers published per year.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eSource: Prepared by the authors.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/13ef7cc8bb6485c6be7d882f.png"},{"id":97460204,"identity":"978dcfe3-d672-466a-94b3-dd8797a5eef0","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":98734,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eDeepfake of Elis Regina in a Volkswagen Kombi advertisement.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eSource: UOL (2023).\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/9513a923b2911aa11b47963e.jpeg"},{"id":97460201,"identity":"efa80fa0-df4e-4358-8f62-682408c936eb","added_by":"auto","created_at":"2025-12-04 15:20:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1207606,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eUsing Blockchain to combat deepfakes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eSource: Adapted from Hasan et al. (2024)\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/971b9a740d5bf1574ca2bbc9.png"},{"id":102294849,"identity":"35e2907d-ca7f-4e00-8405-15ce96f772f9","added_by":"auto","created_at":"2026-02-10 10:00:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1792416,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8118450/v1/8d0c0a9e-230b-455f-ba1b-2381eb885fef.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Ethical and legal implications of using deepfakes in commercials: analysis of the Volkswagen Brazil case and the possibilities of blockchain","fulltext":[{"header":"1. INTRODUCTION","content":"\u003cp\u003eDeepfake is an AI-based technology that allows for the creation or manipulation of video or audio content in a highly realistic way. Using machine learning algorithms, deepfakes can replace one person\u0026apos;s face with another\u0026apos;s, alter facial expressions, or even create videos of people saying or doing things they never actually did (George \u0026amp; George, 2023). The increasingly sophisticated quality of these simulations makes it difficult for the public to distinguish between real and manipulated content, raising serious ethical and legal questions (Chesney \u0026amp; Citron, 2019; Rini, 2020; Rana et al., 2022).\u003c/p\u003e\n\u003cp\u003eThe issue of deepfakes has generated significant interest among marketing professionals and the general population. Despite their relevance to marketing theory and practice, deepfakes\u0026mdash;and the opportunities for benefit or harm they provide\u0026mdash;are poorly understood (Whittaker, Letheren \u0026amp; Mulcahy, 2021). Mazis (1997), Petty (2005), and Celsi (2023) have already highlighted the need for effective public policies to regulate marketing activities, and this is even more pertinent in the context of emerging technologies such as deepfakes.\u003c/p\u003e\n\u003cp\u003eIn the first episode, \u0026quot;Be Right Back,\u0026quot; of the second season of the Netflix series Black Mirror, produced in 2013, the themes of grief, digital inheritance, and the ethical use of Artificial Intelligence (AI) and robotics are explored. Martha, the protagonist, loses her boyfriend Ash in a car accident and is enrolled by a friend in a service that creates a digital replica of Ash using AI. Initially reluctant, Martha interacts more with the replica, which evolves into a robotic form. The story highlights the limitations of technology: the replica, although similar, lacks some of Ash\u0026apos;s characteristics and memories, revealing a less assertive side of the digitization of human identity (InternetLab, 2017; Lacerda \u0026amp; Mattos, 2020).\u003c/p\u003e\n\u003cp\u003eOn July 4, 2023, Volkswagen Brazil launched a commercial to celebrate the brand\u0026apos;s 70th anniversary in the country, featuring a virtual simulation of the legendary Brazilian popular music singer Elis Regina. The ad recreates an interaction between Elis Regina (who died tragically in 1982) and her daughter, Maria Rita, using Artificial Intelligence (AI). In the commercial, while Maria Rita drives the 100% electric version of the Volkswagen Kombi, Elis drives the classic model, providing a moving duet between mother and daughter (The Drum, 2023; Shoot Online, 2023; DataSciConnect, 2023). As seen in the Volkswagen commercial with Elis Regina and in the Black Mirror episode \u0026quot;Be Right Back,\u0026quot; the digital recreation of deceased individuals raises ethical and emotional questions about AI. Both examples show how technology can revisit past relationships and highlight the debate about the emotional and ethical implications of digital recreation using AI and deepfakes.\u003c/p\u003e\n\u003cp\u003eThe advancement of deepfakes has brought serious challenges, as exemplified in the 2024 elections in Brazil. The Brazilian Superior Electoral Court (TSE) faced the spread of manipulated videos that harmed candidates, such as S\u0026atilde;o Paulo mayoral candidate Tabata Amaral, who was the target of false content of a sexual nature, and responded by regulating the use of AI in election campaigns (Superior Electoral Court, 2024; UOL Not\u0026iacute;cias, 2024). In addition, Starlink faced fraudulent ads that used deepfakes to clone the voices of politicians, such as the Brazilian federal deputy Nikolas Ferreira, promoting false Internet offers (Boatos.org, 2024). These cases highlight the need for regulatory, legal, and technological measures to mitigate the risks posed by malicious AI use and protect public and commercial integrity.\u003c/p\u003e\n\u003cp\u003eOn July 10, 2023, the Brazilian National Council for Advertising Self-Regulation (CONAR) initiated an ethical proceeding to investigate Volkswagen\u0026apos;s advertisement that used deepfake and AI techniques to recreate singer Elis Regina. The proceeding was opened following consumer complaints questioning the ethics and the potential for confusion between fiction and reality, especially for children and adolescents.\u003c/p\u003e\n\u003cp\u003eThe creation of a commercial that uses AI to recreate a deceased person, as in the case of Elis Regina, raises ethical and social issues beyond simple product or brand promotion. This practice raises crucial questions about consent, image rights, and cultural impact, which need to be rigorously examined (Harris \u0026amp; Carman, 1983). In addition, there is a significant externality related to the potential for confusion between fiction and reality, especially among more vulnerable audiences such as children, adolescents, and the elderly (Turkle, 1997).\u003c/p\u003e\n\u003cp\u003eThis confusion can have psychological and social implications, such as distorting perceptions of the recreated person and their legacy. Hence, the use of advanced technologies in marketing strategies requires in-depth ethical debate and possibly the intervention of public policies to regulate such practices (Mittelstadt et al., 2016; Naffi et al., 2023; Kumar et al., 2024). The importance of effective public policies to control marketing activities, especially regarding emerging technologies, has been discussed in previous studies (Mazis et al., 1997; Petty \u0026amp; Celsi, 2023; Kshetri et al., 2023) that highlighted the need for careful regulation to protect consumers\u0026apos; and society\u0026apos;s interests amid technological advances.\u003c/p\u003e\n\u003cp\u003eThe central issue of this study arises from the gap between technological advancement and the ethical and legal regulations that can mitigate the negative impacts of these practices. The recreation of deceased individuals through AI, especially for commercial purposes, raises important questions: to what extent is it ethical to recreate someone without their explicit consent digitally? What are the social and cultural consequences of this practice? How can we ensure that the public and especially vulnerable groups such as children and adolescents understand the fiction involved? How can we regulate such practices to balance technological advances and respect for human rights?\u003c/p\u003e\n\u003cp\u003eThis paper has potential contributions in the domains of digital marketing, political marketing and ethical challenges. In the social sphere, it seeks to deepen understanding of the emotional and cultural impacts of using deepfakes and AI to recreate public figures, thereby raising awareness of the issue. Academically and theoretically, the paper addresses the interdisciplinary debate involving technology, ethics, and communication, offering a critical overview of the externalities of digital marketing and proposing innovative solutions based on emerging technologies, such as blockchain. From a marketing perspective, the Volkswagen case analysis serves as a concrete example for brands that wish to explore advanced technologies responsibly and ethically, offering practical guidelines to avoid negative impacts on public perception. For society, the paper contributes by proposing ways to develop public policies that regulate the use of AI and deepfakes, drawing on experiences such as the AI Act approved by the European Parliament, thereby strengthening the legal framework necessary to address the implications of these technologies.\u003c/p\u003e\n\u003cp\u003eBy investigating the externalities of digital marketing and proposing technological and regulatory solutions, this paper not only analyzes the Volkswagen case but also contributes to the understanding of the role of ethics and innovation in a scenario increasingly dominated by AI. Thus, the paper seeks to foster a qualified, interdisciplinary debate on the limits and possibilities of deepfakes in commercial contexts, promoting a balance between technological advancement and social responsibility.\u003c/p\u003e\n\u003cp\u003eBeyond regulatory issues, this paper advocates using blockchain technology as a potential solution to the problem. Dixon (2024) and Seneviratne (2022) highlight the potential of blockchain technology in combating the spread of deepfakes. Dixon (2024) emphasizes that the immutable and transparent record of blockchain is fundamental to the authentication and integrity of multimedia content, facilitating source identification and combating fake videos. Additionally, Seneviratne (2022) outlines the importance of a decentralized and cryptographic system provided by blockchain. When combined with AI techniques, blockchain could verify the authenticity of content and create an integrated approach to tackling false information and deepfakes across different platforms.\u003c/p\u003e\n\u003cp\u003eFollowing the perspective of Motta (2022), this paper proposes solutions with a focus on extrapolation, identifying existing approaches to help reduce the impact of the emerging problem of deepfakes and AI-related spread. The focus for the solution lies in deepening the regulatory debate and using emerging technologies, such as blockchain.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;Beyond this introduction, the paper is structured as follows: background discussing the subjects of deepfake and its regulations and impacts on vulnerable populations; a methodology section detailing the literature review procedures and the documentary analysis; a data analysis section presenting the case of the Brazilian singer deepfake commercial; discussion about the possibilities of using blockchain to mitigate these issues, and, the conclusions.\u003c/p\u003e"},{"header":"2. BACKGROUND","content":"\u003cp\u003eThe Background section is structured into interconnected subsections that progress from the technological and ethical implications of AI and deepfakes (2.1) to the regulatory frameworks addressing these technologies (2.2) and finally to the social impacts on vulnerable populations (2.3). This organization establishes a coherent analytical basis that integrates technological, legal, and societal dimensions, supporting a comprehensive understanding of the deepfake phenomenon within contemporary debates on ethics, governance, and digital transformation.\u003c/p\u003e\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Impacts of AI and deepfakes\u003c/h2\u003e\u003cp\u003eArtificial Intelligence (AI) and deepfake technology have their roots in progressive advances in machine learning and digital manipulation. Initially, deep learning algorithms were applied to recognize patterns in massive datasets (Goodfellow et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), enabling the creation of highly realistic content. Deepfakes, specifically, began to be widely discussed in 2017, when generative adversarial networks were used to create videos and images that even the most experienced analysts could not distinguish from real content (Kietzmann et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Henceforth, technological advances have expanded their applications, from interactive entertainment to malicious uses, such as the dissemination of misinformation or political manipulation (Chesney \u0026amp; Citron, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThese technological advances have brought benefits in fields such as marketing and education, but they have also raised ethical challenges. For example, the use of deepfakes in advertising, such as recreating singer Elis Regina in a Volkswagen commercial, highlights how technology can affect both consumer engagement and trust in digital content (George \u0026amp; George, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe ethical analysis of the use of Artificial Intelligence (AI) to recreate deceased people is a complex issue that begins with the dilemma of consent. since the recreated person cannot give their own consent (Faden, Beauchamp, and Kass, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). This matter raises complex questions about who has the right to give this consent in their place, whether it be the family, the heirs, or a legal entity. This issue extends to image rights, which are protected by law. However, laws can vary from country to country and may not be up to date to address the specific challenges presented by AI (Lessig, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). This context creates a legal vacuum that could allow for the misuse of a person's image, leading to possible legal action.\u003c/p\u003e\u003cp\u003eIn addition to legal issues, there are also considerations regarding the social and cultural impact of recreating public figures. Post (2000) and Naffi et al. (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) argue that this can have a significant impact on society, possibly distorting or trivializing the legacy of the recreated person. This is especially concerning when the figure in question has an important cultural or historical legacy.\u003c/p\u003e\u003cp\u003eThe issue becomes even more complex when considering the risk of confusion between fiction and reality. Turkle (\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e1997\u003c/span\u003e) points out that this risk is exceptionally high among children, adolescents, and the elderly, who may not fully understand that the person has been digitally recreated. Naffi et al. (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) reiterate that this can have psychological and social implications, including distorting perceptions of the recreated person and their legacy.\u003c/p\u003e\u003cp\u003eThe increasing simplicity involved in creating convincing deepfakes, combined with our increasingly digitally documented lives, may increase the potential for deepfakes to be used for malicious purposes, such as misinformation or fake news, blackmail, intimidation, sabotage, harassment, defamation, revenge pornography, identity theft, and bullying (Whittaker et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). On the other hand, through deepfakes, it is possible to create dynamic advertisements for individual consumers\u0026mdash;the ultimate in personalization. Evidence suggests that advertising personalization is positively related to ad persuasiveness and perceived benefit (Ham, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Thus, increased personalization potential may lead to greater desired consumer (re)action (Kietzmanna, Millsb \u0026amp; Planggerc, 2021).\u003c/p\u003e\u003cp\u003eGiven this complex context, Mittelstadt et al. (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) advocate the creation of strict guidelines and regulations for the use of AI to recreate deceased people. These guidelines could include explicit consent requirements, clear disclosure that the person has been digitally recreated, and possibly restrictions on how and where these recreations can be used. In addition to the ethical issues that require regulation, Celsi (2023) outlines the difficulty posed by the rapid AI development and the public authorities' ability around the world to keep up with the technology through applicable legislation.\u003c/p\u003e\u003cp\u003eThe use of deepfake technology, which manipulates videos and audio to create realistic representations of people saying or doing things they did not, presents significant challenges and opportunities from a macromarketing perspective. As a field that studies the interactions between marketing and society, macromarketing needs to consider the ethical, social, and legal implications of this emerging technology. As a meta-theory, Macromarketing analyzes how marketing systems interact with society and reflects on large-scale economic, social, and cultural impacts (Shapiro, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Pittz, Steiner \u0026amp; Pennington, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In the context of AI and deepfakes, macromarketing offers a critical lens for understanding how these emerging technologies shape consumer behavior, cultural dynamics, and ethical issues. By enabling hyper-realistic representations, deepfakes not only redefine advertising practices but also introduce significant risks. Emerging technologies often generate unforeseen negative externalities, which require robust regulations to protect consumers and social balance (Harris \u0026amp; Carman, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e1983\u003c/span\u003e; Hermann, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn the case of recreating public or cultural figures, questions arise about the role of these representations in shaping collective memory and the emotional impact on different audiences. Thus, understanding the implications of using deepfakes through this theoretical perspective is essential to align market practices with ethical and social values. First, the use of deepfakes in marketing can offer new forms of consumer engagement and advertising strategies (Kietzmann, Mills, \u0026amp; Plangger, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). However, this approach comes with substantial ethical risks, such as the spread of misinformation and the violation of image rights and consent, as already discussed by Harris and Carman (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e1983\u003c/span\u003e) in the context of marketing regulation. The ability to create realistic deepfakes can make it difficult for consumers to distinguish between reality and manipulation, exacerbating trust and credibility issues in the relationship between brands and consumers (George \u0026amp; George, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eFurthermore, the use of deepfakes may have broader social and cultural implications. For example, the recreation of public or cultural figures can affect public perception and their legacy, a topic Hutt, Mokwa, and Shapiro (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1986\u003c/span\u003e) explore in their discussion of the \"parallel political market\" in marketing. This matter also raises questions about representation and identity, mainly when deepfakes are used to alter characteristics such as race, gender, or age.\u003c/p\u003e\u003cp\u003eAlthough deepfake technology has impressive applications, its effects on vulnerable populations require special attention. Children, adolescents, and the elderly, for example, have greater difficulty distinguishing between fiction and reality (Turkle, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). This inability to discern can have psychological impacts, such as the idealization or distortion of digitally recreated public figures, especially when they involve sensitive topics such as grief or nostalgia (Naffi et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eA relevant example is the use of deepfakes for political manipulation, which can exploit vulnerable populations' lack of access to reliable information, thereby increasing the risk of misinformation. This analysis reinforces the need for public policies that account for the impact of these technologies on vulnerable audiences, promoting digital literacy and the development of critical skills.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Deepfakes regulation\u003c/h2\u003e\u003cp\u003eThe regulation of deepfake technology is still a developing field, with different jurisdictions adopting different approaches to address the ethical and legal challenges it presents. In the United States, for example, initiatives such as the Deepfake Accountability Act (a federal bill that criminalizes the production of deepfakes without identifying that AI generates the content), enacted in 2023, seek to establish a specific legal framework for this technology, requiring the disclosure of manipulated videos or audios and holding creators liable for any damage caused (Chesney \u0026amp; Citron, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In the European Union, the approach has been more comprehensive, with the consideration of regulatory frameworks that address various applications of Artificial Intelligence, including deepfakes (European Commission, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), through the Digital Services Act published in 2022 (European Commission, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and the Artificial Intelligence Act (AI Act) signed in early 2024 (European Commission, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe European Union's AI Act represents a comprehensive regulatory framework that aims to ensure safety, ethics, and transparency in the AI development and use in member states. By classifying AI systems into risk categories, the European Act prohibits those posing an unacceptable risk and imposes strict obligations on high-risk systems. Although it does not explicitly mention deepfakes, the AI Act addresses manipulative and deceptive techniques that may cover the harmful uses of deepfakes. For high-risk systems and general-purpose AI models, detailed technical documentation, data governance, and transparency measures are required to ensure compliance with ethical and legal guidelines (Future of Life Institute, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThis evolving regulatory landscape makes cases such as Volkswagen's particularly instructive. In Brazil, where government regulation of deepfakes is still in its infancy, self-regulatory bodies such as CONAR play a crucial role (CONAR, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). However, between 2010 and 2020, there was a high number of complaints to CONAR related to advertising aimed at children and adolescents. Most of these consumer complaints were dismissed (Matos et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eSix legislative proposals in Brazil address this issue, but as all proposals require approval by both houses of the National Congress, none have been enacted to date. The first initiative was in 2019 and came from the Federal Senate, with Bill No. 5,051/2019, which aims to regulate the AI use in Brazil (Federal Senate, 2024). Also in 2019, the same author of the previous proposal presented Bill No. 5,691/2019, which aimed to establish the National Artificial Intelligence Policy (Federal Senate, 2024). The following year, Bill No. 21/2020 was proposed in the Chamber of Deputies \u0026mdash;the so-called Legal Framework for AI \u0026mdash;with the intention of establishing principles, rights, duties, and governance instruments for the use of AI in Brazil (Federal Senate, 2024). Even without the approval of any of the previous proposals, in 2021, the Senate received a new proposal, Bill No. 872/2021, whose objective is to provide ethical frameworks and guidelines for the development and use of AI in Brazil (Federal Senate, 2024). In 2023, two new motions were presented, one in the Chamber of Deputies, Bill No. 759/2023, focused on regulating AI systems (Chamber of Deputies, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and in the Federal Senate, Bill No. 2,338/2023, focused on AI usage (Federal Senate, 2024).\u003c/p\u003e\u003cp\u003eAlthough providing important guidelines on AI, the Brazilian Bill 2338/2023 contains articles that, lack greater precision and specificity, resulting in subjectivity that could compromise the practical application of the regulations (Chamber of Deputies, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The use of vague and adjectival terms leads to inconsistent interpretations, especially of central concepts such as model, impact, and associated risk. These elements require clear, objective definitions for the legislation to be practical and applicable across diverse scenarios.\u003c/p\u003e\u003cp\u003eOne of the main problems with the text is the absence of objective criteria for classifying an AI system as \"high risk.\" This flaw creates ambiguity and can lead to uneven interpretations. Based on international frameworks, such as the European Union's AI Act, and the discussions of the first OBIA Seminar (Brazilian Observatory of Artificial Intelligence about algorithmic auditing (Meira, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), the classification of high-risk systems should consider elements such as probability of failure, severity of impact, and scope of automated decisions. If incorporated into the project, these criteria would provide a more solid basis for identifying high-risk systems and targeting specific regulations.\u003c/p\u003e\u003cp\u003eAnother point of criticism is the lack of attention to the regulation of emerging technologies, such as deepfakes. Despite their obvious risks, the bills under consideration address deepfakes only in the electoral context, as in the Brazilian law project 1.002/2023 (Chamber of Deputies, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This approach disregards the dangers of this technology in other areas, including large-scale disinformation, financial fraud, and emotional manipulation. To correct this gap, it would be necessary to include specific regulations addressing deepfakes in commercial and advertising contexts, as well as preventive measures, such as mandatory disclaimers stating that certain content was generated or manipulated by AI and the implementation of tools that guarantee the authenticity of digital content (Nightingale \u0026amp; Wade, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Dixon, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Dixon, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAlthough the Brazilian Bill 2338/2023 emphasizes fundamental principles such as transparency, security, privacy protection, and respect for ethics and human rights, it fails to detail how these values will be operationalized. For example, while highlighting the promotion of sustainable and inclusive development, it does not present specific measures to protect vulnerable populations from discriminatory or manipulative technologies. One solution would be to detail oversight mechanisms, such as the creation of independent committees, and to establish clear technical standards based on international standards, such as ISO/IEC 23894, which addresses AI governance (ISO/IEC, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). About governance tools and representative data, Article 18 of Brazilian Bill 2338/2023 requires adequate training, validation, and testing of data, but does not define the objective criteria that such data must meet. Adequate data must be bias-free, up-to-date, and representative of the context in which the system will be applied (Hwang et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In addition, the automatic recording tools mentioned in the text should include standards for accuracy and robustness, ensuring the traceability and auditability of automated decisions. Overall, Brazilian law project 2338/2023 reflects a legitimate concern with the regulation of AI in Brazil, but it has significant shortcomings. The lack of conceptual clarity, disregard for emerging technologies such as deepfakes, and the guidelines' generality weaken their applicability. Incorporating objective criteria for classifying high-risk systems, expanding the discussion of deepfakes, adopting internationally recognized technical standards, and strengthening oversight mechanisms would be essential steps to improve the legal text and ensure it aligns with society's needs and technological advances. Hence, the issue of deepfake regulation is complex and requires debate and further study, involving interaction between government laws and self-regulatory mechanisms (Whittaker, Letheren \u0026amp; Mulcahy, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). As technology continues to evolve, a more coordinated and perhaps international approach to addressing ethical challenges is likely to emerge.\u003c/p\u003e\u003cp\u003eThe impacts of deepfakes undoubtedly warrant further regulatory consideration. However, from the perspectives of literacy and understanding on the part of companies and vulnerable groups, it is essential to intensify research demonstrating how educational techniques and transformative research can be applied to raise consumer awareness of the threats.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Impact of deepfakes on vulnerable populations\u003c/h2\u003e\u003cp\u003eThe impact of technologies such as deepfakes on vulnerable populations is a critical area that requires attention from regulators, researchers, and educational institutions. Although all segments of the population are subject to the risks of misinformation and digital manipulation, groups such as children, adolescents, the elderly, and individuals with less access to digital education are particularly susceptible. Studies indicate that these populations have greater difficulty distinguishing between real and manipulated content, mainly due to their reliance on unverified sources and lack of digital literacy (Hwang et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Nightingale \u0026amp; Wade, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Naffi et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe difficulty in recognizing deepfakes not only poses a risk to the spread of misinformation but can also cause significant psychological and social impacts. Children and the elderly, for example, are especially prone to internalizing manipulated information as accurate, which can lead to widespread distrust of legitimate sources of information (Nightingale \u0026amp; Wade, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). These effects can be amplified in emotionally charged contexts, such as the manipulation of political discourse or the use of deepfakes for fraud and scams. This demonstrates the urgent need for educational and technological measures to mitigate such impacts.\u003c/p\u003e\u003cp\u003eOne of the most effective strategies for reducing the impact of deepfakes on vulnerable populations is implementing digital literacy programs. Educational programs aimed at children, adolescents, and the elderly, focused on teaching the basic principles of digital manipulation and the characteristics that can expose a deepfake, are essential. Examples of such initiatives are already being tested in educational institutions worldwide (Hwang et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAlthough education is crucial, it must be complemented by measures implemented by digital platforms. Companies such as Meta and X (formerly Twitter) need to adopt deepfake-detection technologies to alert users to potential manipulations. In addition, specific regulations requiring clear disclaimers for AI-generated content would be a significant step toward ensuring transparency.\u003c/p\u003e\u003cp\u003eIn addition to theoretical and technological discussions, there is an urgent need for an integrated approach to mitigate the impacts of deepfakes on vulnerable populations. Education, regulation, and technological solutions must be implemented simultaneously, involving governments, technology companies, educational institutions, and civil society. The role of digital platforms in detecting and flagging manipulated content, combined with digital literacy and blockchain, represents a promising way to protect more vulnerable populations.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. METHODOLOGY","content":"\u003cp\u003eThis study adopted the case study approach (Yin, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), systematically investigating contemporary phenomena within their real-life contexts. Initially, the objectives of the literature review were established to outline the research scope within the context of deepfakes, focusing specifically on public policy and technological innovation. The subsequent phase involved searching databases of recent papers and online materials, such as legislation and media coverage, to ensure the research included relevant studies with diverse perspectives and approaches to the topic.\u003c/p\u003e\u003cp\u003eMultiple data sources were used to increase the reliability of the results (Eisenhardt, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1989\u003c/span\u003e). Documentary research constituted a significant part of the data collection, involving detailed analysis of online materials, including proposed legislation and regulations, academic papers, and media coverage of the application of deepfakes in advertising. This exploration was expanded through a literature review across recognized academic databases, including Google Scholar, EBSCO, and the AI tools Consensus and Typeset. The following keywords were used separately: deepfakes AND \"artificial intelligence\"; deepfakes AND regulation; and deepfakes AND blockchain for the period from 2016 to March 2024. A total of 32 papers were identified in the databases searched. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows that 2023 had the highest number of publications (10) related to the search terms, likely reflecting the growth of interest in AI, particularly after the launch of OpenAI's ChatGPT in late 2022. The objective of the documentary research and literature review was to ensure a comprehensive investigation of the ethical and legal debates surrounding deepfakes, encompassing contributions from various disciplines to reflect the topic's interdisciplinary nature (Yin, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe search and filtering stage consisted of applying the terms established in the databases, followed by selecting papers based on predefined criteria focused on the recent publication period, availability of the complete text, and direct relevance to the impacts of deepfakes. For data extraction, information such as the authors' names, year of publication, study objectives, methodologies used, and main conclusions of each paper were collected. This data collection was instrumental for the subsequent analysis.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSource: Prepared by the authors.\u003c/p\u003e\u003cp\u003eThe AI-based qualitative data analysis tool typeset.io was used to assist in organizing and analyzing the vast amount of material collected, contributing to the integrity and systematization of the literature review results. The critical evaluation of the methodological quality of the selected papers was a crucial step aimed at identifying potential biases and ensuring the reliability of the sources. In addition, efforts were made to mitigate publication bias by examining gray literature and documents in different languages.\u003c/p\u003e\u003cp\u003eThrough this descriptive and systematic methodology, it was possible to develop a comprehensive picture of the academic debates and political discussions on deepfakes, establishing a robust basis for future research and policy development.\u003c/p\u003e"},{"header":"4. DATA ANALYSIS","content":"\u003cp\u003eThe Data Analysis section is divided into three interrelated subsections that collectively examine the empirical dimension of the research. Subsection \u003cspan refid=\"Sec7\" class=\"InternalRef\"\u003e4.1\u003c/span\u003e presents the Volkswagen case, describing the use of deepfake technology to recreate singer Elis Regina in a 2023 advertisement and the ensuing ethical controversy evaluated by CONAR (Brazilian Council for Advertising Self-Regulation). Subsection \u003cspan refid=\"Sec8\" class=\"InternalRef\"\u003e4.2\u003c/span\u003e deepens the analysis by examining CONAR\u0026rsquo;s decision, highlighting tensions between innovation, ethical responsibility, and the protection of vulnerable audiences, while also addressing the absence of clear guidelines on the disclosure of AI-generated content. Finally, subsection \u003cspan refid=\"Sec9\" class=\"InternalRef\"\u003e4.3\u003c/span\u003e broadens the discussion by analyzing other recent deepfake cases\u0026mdash;including political and corporate incidents\u0026mdash;that demonstrate the global diffusion of such technologies and the urgent need for more robust regulatory, educational, and technological mechanisms. Together, these subsections construct an analytical framework that integrates legal, ethical, and social perspectives, elucidating how deepfakes challenge the boundaries between creativity, manipulation, and governance in contemporary digital communication.\u003c/p\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Case description\u003c/h2\u003e\u003cp\u003eOn July 4, 2023, Volkswagen Brazil launched an advertisement using deepfake technology to present a duet between the late Brazilian singer Elis Regina and her daughter, Maria Rita, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. However, this marketing strategy was met with rejection, leading to its suspension by CONAR due to ethical concerns related to the image use of a deceased person with the family authorization but without the consent of the rights holder (UOL, 2023; G1, 2023; Folha de S\u0026atilde;o Paulo, 2023; Estad\u0026atilde;o, 2023).\u003c/p\u003e\u003cp\u003eRegarding Volkswagen's advertising campaign in Brazil, the issue raised by the complaint (Estad\u0026atilde;o, 2023) highlights the challenge of distinguishing between fiction and reality in advertising and the possibility that children and adolescents may not readily perceive this distinction. The use of AI to create content that mimics reality with high precision \u0026mdash;known as deepfakes \u0026mdash;highlights the ability of these representations to influence the perceptions of young audiences. Due to their still-developing cognitive abilities, such audiences may have difficulty discerning the artificial nature of such content, leading to possible misunderstandings or false beliefs.\u003c/p\u003e\u003cp\u003eCONAR chose to dismiss the ethical complaint against Volkswagen's advertising campaign, which used deepfake technology to recreate the image of the late singer Elis Regina. This decision was made on August 22, 2023 (Migalhas, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Correio Braziliense, 2023; Carta Capital, 2023). Despite the controversy, Volkswagen celebrated the dismissal of the case, and the campaign featuring Elis Regina achieved record views (Livedune, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eRegarding the issue of image rights, there was unanimous agreement that there was no disrespect to Elis Regina, since her heirs consented to the use of the singer's image, who appeared in the advertisement singing, something she did in life. About image rights, this unanimous decision established that CONAR and Elis Regina's heirs agreed that the recreation was carried out respectfully and ethically. Regarding the impact on vulnerable audiences, especially the need to explicitly inform them about the use of AI in advertisements, the vote was not unanimous. The majority of votes (13 versus 7) found that the AI technique used was evident in the advertising campaign and that no further clarification was needed for the public regarding the use of this technology to recreate Elis Regina's image (Migalhas, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eProduced by the AlmapBBDO agency, the campaign was considered by Volkswagen to be one of its greatest marketing successes. The campaign not only celebrated an important milestone for the company but also explored an emerging technological innovation, showcasing the singer digitally recreated using deepfake technology, which generated considerable media and public discussion and attention (Automotive Business, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAs for the campaign results, Volkswagen announced that the commercial achieved a record number of views, highlighting its positive impact on the public. The campaign resonated with the public, divided opinion, and demonstrated the power of technological innovation to create emotionally engaging advertising content (Livedune, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Volkswagen's defense of the campaign and disclosure of the figures emphasized how the company valued innovation and creativity, despite the controversies and ethical questions raised by the use of deepfake technology to recreate the image of a deceased figure.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Analysis of the decision of the advertising council\u003c/h2\u003e\u003cp\u003eCONAR's (Brazilian Council for Advertising Self-Regulation) decision on Volkswagen's advertising campaign reflects several complexities inherent in emerging advertising and technology practices. In the ethical and regulatory sphere, the decision first addresses image rights. According to Faden, Beauchamp, and Kass (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), image rights are a fundamental ethical principle, particularly in cases of digital recreation of deceased individuals. CONAR's decision, which found no violation of Elis Regina's image rights with her heirs' consent, is aligned with the guidelines for informed consent and respect for image rights (Lessig, 1999).\u003c/p\u003e\u003cp\u003eHowever, analysis of the decision reveals significant gaps in the treatment of vulnerable audiences and the need for greater transparency. Evidence suggests that vulnerable groups, such as children and the elderly, may have difficulty distinguishing reality from the digital manipulation presented in the commercial, especially given the sophistication of deepfake technologies (Turkle, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Scherer \u0026amp; Pennycook, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). This cognitive limitation, combined with the emotional impact of seeing public figures recreated so realistically, raises concerns about possible psychological and social effects. Studies indicate that the confusion generated by hyper-realistic digital content can distort perceptions about the legacy of the recreated person, exacerbating the risk of unrealistic idealizations or even trivializing their memory (Naffi et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eFurthermore, the debate at CONAR on the impact of advertising on vulnerable audiences raises the question of the duty to explicitly inform about the use of AI in the production of the ad. The decision that the technique's use was evident was not unanimous, highlighting the vulnerability of certain groups to misinformation \u0026mdash;an issue that intensifies with age and trust in unverified sources (Scherer \u0026amp; Pennycook, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn this context, CONAR's decision not to explicitly require a disclaimer stating that the campaign used AI and deepfake techniques reveals a missed opportunity to set a more robust ethical precedent. The use of clear and visible disclaimers could not only educate the public about the technological nature of the content but also protect vulnerable groups from confusion and promote greater transparency in advertising practices. This preventive measure is widely advocated in the literature as an essential approach to mitigating the risks associated with misinformation and digital manipulation (Mirsky \u0026amp; Lee, 2020).\u003c/p\u003e\u003cp\u003eAdopting mandatory guidelines to include disclaimers in campaigns that use advanced technologies, such as AI, could better align advertising practices with the principles of self-regulation and ethical responsibility. Furthermore, such an approach would align with international efforts to regulate emerging technologies, such as those proposed in the European Union's AI Act, which emphasize transparency and consumer protection in vulnerable situations.\u003c/p\u003e\u003cp\u003eCONAR's decision highlights the importance of moving forward with building a more comprehensive regulatory framework that considers not only the legal and ethical aspects but also the social and cultural impacts of technologies such as AI. The inclusion of measures such as the mandatory use of disclaimers could serve as a model for proactive regulation, promoting more responsible advertising and adequately protecting consumers in an increasingly complex digital environment. In this context, CONAR's decision underscores the importance of advertising self-regulatory bodies in a regulatory landscape where government regulations are still under development. However, the decision suggests the need for a more holistic regulatory framework that incorporates the social and cultural impacts of technologies such as artificial intelligence, going beyond mere legal and ethical considerations (Post, 2000).\u003c/p\u003e\u003cp\u003eIt is essential to note that this overview of CONAR's decision reflects on how organizations should navigate the use of emerging technologies, balancing innovation and creativity with ethical and legal responsibility. The ease with which this technology can be used for unethical and malicious purposes, such as spreading misinformation and falsifying identities, has been a growing concern (Mirsky \u0026amp; Lee, 2020; Grewal, Guha, \u0026amp; Becker, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Analysis of other deepfake cases and recent events\u003c/h2\u003e\u003cp\u003eSocial media has often been the scene of cases that illustrate the urgency of more robust regulation. Notable examples include the fake images of the Pope wearing a stylish white coat and the fake arrest of Donald Trump (Metropoles, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), both in 2023. In early 2024, an employee of a multinational company transferred \u003cspan\u003e$\u003c/span\u003e25\u0026nbsp;million to criminals after a video call with a Chief Financial Officer produced by a deepfake (CNN, 2024). These incidents highlight the need for a more comprehensive regulatory approach to mitigate the risks posed by irresponsible use of technologies such as deepfakes.\u003c/p\u003e\u003cp\u003eThe proliferation of AI-generated images depicting fake events raises significant concerns about people's ability to discern their authenticity. Not everyone has the technical knowledge or attention to detail necessary to identify the subtleties that distinguish a real image from a fake one. This poses a serious risk, as misleading images can spread misinformation, influence public opinion, and even manipulate political and social events. The difficulty in verifying image authenticity can lead to the spread of false information, undermining trust in digital content and negatively impacting society (Kietzmann et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn February 2024, OpenAI announced its new AI called Sora. OpenAI's Sora can generate realistic, creative videos from textual instructions. With the ability to create videos up to one minute long, Sora maintains visual fidelity and precisely follows user requests, generating complex scenes that include multiple characters, specific movements, and accurate details of scenery and objects (OpenAI, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAs of March 2024, the platform was not available to the public because, according to OpenAI (\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), the company was conducting tests and developing tools to detect misleading content, aiming to prevent misinformation, hateful content, and bias. However, Sora does not fall into the category of explainable AI (XAI), which raises questions about its functioning and transparency, as well as the potential risk of generating deepfakes on the platform. XAI refers to an interpretable system that provides a comprehensible explanation of its output and draws attention to the ethics of AI, focusing on how AI systems make decisions, the explanations for those decisions, and how those decisions are communicated to relevant stakeholders (Vainio-Pekka et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe advancement of deepfakes has brought serious challenges, as evidenced in the 2024 elections in Brazil. The Superior Electoral Court (TSE) faced significant problems with digitally manipulated videos, which harmed candidates by spreading false information. In response, the TSE regulated the use of AI in election campaigns, expressly prohibiting the creation and propagation of deepfakes. In addition, it determined that the AI use in campaigns should be accompanied by an explicit warning to promote greater transparency (Superior Electoral Court, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eOne of the most notorious cases involved S\u0026atilde;o Paulo mayoral candidate Tabata Amaral, who was targeted by fake videos. Digitally manipulated images associating her with sexual content were posted on social media to damage her public image. In response, the candidate took legal action through the Electoral Court to hold those involved accountable and combat misinformation during the election period (UOL Not\u0026iacute;cias, 2024).\u003c/p\u003e\u003cp\u003eStarlink also faced challenges related to the use of deepfakes. Scammers created fraudulent ads promoting fake lifetime internet plans, using videos that cloned the voice of Brazilian congressman Nikolas Ferreira. This misleading content tricked consumers into visiting fraudulent websites, compromising the security of their personal and financial data and damaging the company's reputation (Boatos.org, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThese cases highlight the risks of malicious AI use and reinforce the importance of regulatory, legal, and technological measures to combat deepfakes, protecting the integrity of electoral processes and the reputations of individuals and organizations. In this sense, the use of Blockchain may be an appropriate alternative.\u003c/p\u003e\u003c/div\u003e"},{"header":"5. DISCUSSION","content":"\u003cp\u003eConsidering the technological nature of this paper, the discussion section emphasizes the proposition of a solution to the identified issues. Hence, blockchain emerges as a promising technology to mitigate the problems associated with deepfakes. Through its decentralized, immutable structure, blockchain can authenticate and track the origin of digital content, increasing confidence in the material disclosed. Dixon (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) points out that blockchain's transparent record is essential to ensuring media integrity, while Seneviratne (\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) suggests that its integration with AI can offer robust solutions for verifying content authenticity.\u003c/p\u003e\u003cp\u003eIn the context of advertising and the ethical use of AI, blockchain could ensure that digital recreations, as in the Volkswagen case, are appropriately labeled and verified for origin, reducing the risk of confusion between reality and fiction. This technological solution not only addresses ethical issues but also strengthens consumer confidence in AI-based marketing practices.\u003c/p\u003e\u003cp\u003eKnown for its immutable, transparent transaction record, blockchain could be decisive in authenticating the origin and maintaining the integrity of multimedia content. Dixon (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) suggests that blockchains, by providing an unalterable audit trail, help identify the source of digital content \u0026mdash;a crucial element in combating the spread of fake videos.\u003c/p\u003e\u003cp\u003eSeneviratne (\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) also argues that blockchain technology can significantly reduce the prevalence of deepfakes by ensuring the authenticity and integrity of visual media. By leveraging the decentralized nature, encryption capabilities, and tamper-proof features of blockchain, a secure and reliable system can be established to verify the authenticity of content. The author argues that by combining blockchain technology with AI techniques \u0026mdash;such as tamper-proof blockchain and AI models to detect fake news \u0026mdash;a synergistic approach can be developed to combat the proliferation of misleading information and deepfakes across various platforms.\u003c/p\u003e\u003cp\u003eBlockchain-based technologies offer a promising approach to verifying the authenticity of digital media, playing a crucial role in combating the proliferation of deepfakes. Chan et al. (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) highlight blockchain's ability to provide a decentralized architecture that enables the verification of digital content authenticity and the tracking of its historical provenance. This feature is especially pertinent, as deepfakes are becoming increasingly sophisticated and challenging to identify. Additionally, Prajapati et al. (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) suggest using blockchain to authenticate products through linked QR codes, reinforcing trust and authenticity and, consequently, mitigating the circulation of counterfeit products. This approach demonstrates the versatility of blockchain beyond cryptocurrencies, expanding its use to digital security across the broad spectrum.\u003c/p\u003e\u003cp\u003eJing and Murugesan (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) complement this view, proposing that blockchain can effectively reduce the incidence of deepfakes by implementing a trust index model on social media. This model prioritizes data privacy, credibility, and transparency, creating a more secure and reliable digital environment where false information and media manipulation are readily identifiable and discredited.\u003c/p\u003e\u003cp\u003eIn the context of deepfakes, a key aspect is the importance of double-checking information upon receipt, especially given the significant challenge of accurately detecting them. The ability to verify information from a reliable source is crucial, especially in educational processes that aim to protect vulnerable groups, such as children and the elderly. Deepfakes have great potential for negative impact, given their ability to imitate reality compellingly. Automated detection of deepfakes is essential to prevent the spread of false information that can manipulate public opinion, compromise electoral processes, and cause profound social and economic damage (Verdoliva, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In addition, Shahzad et al. (\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) emphasize the importance of combining technological solutions with digital education and awareness to mitigate the harmful effects of this technology.\u003c/p\u003e\u003cp\u003eAnother relevant point in tackling deepfakes is the use of blockchain as a supporting technology. Blockchain offers several improvements that should be explored, especially for certifying the creator's identity and the authenticity of content. This technology can provide greater security through immutable and decentralized records, creating a reliable validation system for digital publications (Sudhakar \u0026amp; Shanthi, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Heidari et al. (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) argue that NFTs (Non-Fungible Tokens), associated with blockchain, can be an effective tool for authenticating content, ensuring origin and credibility through immutable records. Sam Altman, from OpenAI, proposed the WorldID platform, which uses blockchain to ensure individuals' universal identity. This proposal has the potential to reduce the spread of false information by eliminating anonymity, as people who spread fake news would have no way to hide (Singh \u0026amp; Dhiman, 2023).\u003c/p\u003e\u003cp\u003eIn addition, decentralized governance systems can be integrated with technologies such as oracles, allowing users to attest to the authenticity of information, thereby promoting a more resilient ecosystem against misinformation (Gupta et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This integration of blockchain systems with external platforms facilitates the creation of reputation rankings for content and its authors, validating reliability through distributed governance mechanisms. As noted by Singh \u0026amp; Dhiman (2023), decentralization mitigates the harms of a centralized authority with complete control over reputation, thereby enhancing system resilience and reliability.\u003c/p\u003e\u003cp\u003eHasan et al. (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) proposes an innovative solution based on blockchain and NFTs to address the challenges posed by deepfake technology, particularly in the metaverse and digital media. The solution uses NFTs to authenticate and protect the originality of digital content. Digital content creators register on an NFT marketplace (NFTM), where they can create and trade digital assets linked to NFTs. The blockchain ensures the immutability of records and the provenance of data, while smart contracts enable secure, traceable transactions.\u003c/p\u003e\u003cp\u003eIn addition, the proposal integrates oracles to update dynamic NFT metadata, allowing the status of digital assets to be adjusted based on external data (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). To mitigate the high costs of on-chain storage, NFT metadata is stored in a decentralized manner using systems such as IPFS (Interplanetary File System), with only the file hash stored on the blockchain. This model reinforces data security and integrity while ensuring the authenticity and reliability of content through a decentralized reputation system that assigns \"trust badges\" to creators based on their trustworthiness. With this approach, the paper proposes a robust method for addressing the risks associated with deepfakes, ensuring the authenticity, ownership, and originality of digital assets.\u003c/p\u003e\u003cp\u003eA multifaceted approach combining regulation, technology, and educational processes is needed to effectively combat the challenges posed by deepfakes. Shoaib et al. (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) emphasize that public policies aligned with technological solutions, such as blockchain integration, are essential to mitigate social, political, and economic impacts. Complementarily, Rana et al. (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) emphasize that digital education and the development of automated detection tools are indispensable for reducing the impact of these threats in the contemporary world.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSource: Adapted from Hasan et al. (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\u003cp\u003eThe triangulation of these perspectives highlights the applicability of blockchain as a robust tool against misinformation. The synergy between blockchain decentralization, content and product authentication, and the creation of a trust index on social media establishes a theoretical and practical framework for mitigating the challenges posed by deepfakes. Therefore, adopting this emerging technology can help ensure information integrity in the digital age, paving the way for a technological response to the advancement of media manipulation techniques.\u003c/p\u003e"},{"header":"6. CONCLUSION","content":"\u003cp\u003eAlthough efforts are being made to establish ethical and legal guidelines, the Volkswagen case and CONAR's decision indicate that the issue should be discussed further, given the incipient nature of laws and regulations addressing the use of emerging technologies, such as AI in advertising and deepfakes. Discussions on regulation in 2024 should be deepened to address not only legal issues but also the social and cultural impacts of these technologies (Post, 2000; Kietzmann, Mills, \u0026amp; Plangger, 2020; Celsi, 2023).\u003c/p\u003e\u003cp\u003eThe use of emerging technologies, such as AI and deepfakes, represents both an opportunity for innovation and a significant ethical and regulatory challenge. This technological paper sought to investigate the ethical, social, and cultural implications of deepfakes in advertising. The analysis highlighted the gap between technological advancement and regulation that could mitigate the negative impacts of these practices, offering reflections on consent, protection of vulnerable audiences, cultural impacts, and practical solutions, such as blockchain.\u003c/p\u003e\u003cp\u003eIn addition, the paper contributes by addressing the Brazilian and international regulatory landscape, presenting a compilation of the main laws under discussion and the bills currently being debated in the Brazilian Congress. This analysis included a detailed review of existing and proposed regulations governing the use of deepfakes and other emerging technologies across commercial, cultural, and social contexts.\u003c/p\u003e\u003cp\u003eThere is a noticeable gap between rapid technological advancement and the ability of laws to keep pace with these changes. In Brazil, among the bills mentioned, law project 2630/2020, known as the \"Fake News Law,\" stands out. It aims to establish guidelines for the transparency of digital platforms and accountability for false or manipulated content. Another relevant bill is law project 2338/2023, which proposes a regulatory framework for the protection of personal data and privacy, including the use of AI technologies.\u003c/p\u003e\u003cp\u003eThe data presented show that current regulatory decisions, such as those made by CONAR, still struggle to keep pace with technological development. Although the consent of Elis Regina's heirs ensured respect for image rights, the absence of clear guidelines on transparency, such as the requirement for disclaimers informing the use of deepfakes, reveals a missed opportunity to protect consumers, especially the most vulnerable. Recent cases, such as the deepfakes disseminated during the 2024 elections and scams involving voice cloning to promote fake Starlink plans, reinforce the need for more agile and effective regulations.\u003c/p\u003e\u003cp\u003eIn addition, the examples analyzed demonstrate that hyper-realistic digital manipulation can cause confusion between fiction and reality, especially among children, adolescents, and the elderly, who have greater difficulty discerning the nature of the content. This confusion can have emotional and cultural impacts, such as distorting public perception of digitally recreated historical or cultural figures. Therefore, the mandatory inclusion of disclaimers for the use of technologies such as blockchain to authenticate digital content presents a practical and ethical solution to mitigate these risks, strengthening public confidence and promoting more responsible business practices.\u003c/p\u003e\u003cp\u003eThis study has some limitations, including its focus on a single case and its lack of a comprehensive analysis of diverse legal and cultural contexts. Future contributions could include exploring the deepfake phenomenon from other empirical studies, as well as a more in-depth analysis of how legislation should address ethical and social issues related to the use of deepfakes, especially when they involve deceased persons (Chesney \u0026amp; Citron, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Rini, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Kietzmann, Mills, \u0026amp; Plangger, 2020). It would be relevant to investigate how advertising could incorporate clear messages about the use of deepfakes. Issues regarding the protection of vulnerable populations are also crucial (Naffi et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Mechanisms could be developed to help identify deepfakes and promote open debate in society about their impact.\u003c/p\u003e\u003cp\u003eIn addition, using blockchain to verify the authenticity of videos from reputable sources, such as established media outlets or recognized institutions, would introduce a valuable tool for distinguishing between factual and fabricated content. Dixon (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) argues that blockchain can add a layer of security and authentication to digital content, helping restore trust in information shared online. Linking technologies such as blockchain to mandatory transparency practices, such as the use of explicit disclaimers informing the use of deepfakes, reinforces both the credibility of the content and consumer protection. This approach allows companies to innovate responsibly, promoting confidence in the use of advanced technologies.\u003c/p\u003e\u003cp\u003eFuture research could explore how deepfakes impact sectors beyond advertising, such as journalism and entertainment, and how different regulatory approaches can mitigate their risks. In addition, empirical studies could investigate public perception of deepfakes and the effectiveness of solutions such as disclaimers and blockchain in protecting against digital manipulation. Another aspect that deserves investigation is the role of marketing managers' ethical awareness in the use of these technologies. What are the accepted ethical limits in the market? How can companies incorporate governance practices that prioritize transparency and consumer protection? These issues are fundamental to building a more responsible market that is aligned with social values.\u003c/p\u003e\u003cp\u003eFinally, bills on the subject could benefit from a broader discussion involving not only Congress but also civil society to promote greater awareness of the impacts of this technology. The use of deepfakes and AI in marketing poses significant challenges that require a collective effort among companies, regulatory agencies, governments, and other actors. Self-regulatory bodies, such as CONAR, play an important role in maintaining ethical standards, but the need for more comprehensive and up-to-date government laws and regulations is undeniable. The speed of technological innovation requires an equally agile regulatory response that considers the multiple impacts of these technologies on different segments of society. The future of these AI-based technologies will depend, in large part, on the ability to regulate and use them in ways that respect individual rights and promote collective benefits. In short, the future of digital manipulation technologies depends on our ability to balance innovation with ethical responsibility, protecting the rights of all citizens, especially the most vulnerable.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eD.D.G. wrote the main manuscript text, R.B.C. reviewed and wrote some paragraphs of background, introduction, and discussion, and R.S.L. and H.C.N. reviewed the manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003cp\u003eAutomotive Business. (2023). Volkswagen defends campaign with Elis Regina. \u003cem\u003eAutomotive Business\u003c/em\u003e. https://www.automotivebusiness.com.br\u003c/p\u003e\n\u003cp\u003eBBC. (2023, March 23). Fake Trump arrest photos: How to spot an AI-generated image. \u003cem\u003eBBC News\u003c/em\u003e. https://www.bbc.com/news/world-us-canada-65069316\u003c/p\u003e\n\u003cp\u003eBoatos.org. (2024). A scam uses Nikolas Ferreira's voice, dubbed with AI, to talk about Starlink. \u003cem\u003eBoatos.org\u003c/em\u003e. https://www.boatos.org/tecnologia/nikolas-ferreira-anunciou-que-elon-musk-vai-dar-starlink-vitalicia-por-r-184.html\u003c/p\u003e\n\u003cp\u003eChamber of Deputies. (2023). \u003cem\u003eThe proposal regulates the use of artificial intelligence\u003c/em\u003e. https://www.camara.leg.br/noticias/968967-proposta-regulamenta-utilizacao-da-inteligencia-artificial\u003c/p\u003e\n\u003cp\u003eChamber of Deputies. (2024). \u003cem\u003eBill No. 1,002/2023\u003c/em\u003e. https://www.camara.leg.br/propostas-legislativas/2350539\u003c/p\u003e\n\u003cp\u003eChamber of Deputies. (2024). \u003cem\u003eBill No. 759/2023\u003c/em\u003e. https://www.camara.leg.br/propostas-legislativas/2349685\u003c/p\u003e\n\u003cp\u003eChan, C. C. K., Kumar, V., Delaney, S., \u0026amp; Gochoo, M. (2020). Combating deepfakes: Multi-LSTM and blockchain as proof of authenticity for digital media. In \u003cem\u003e2020 IEEE/ITU International Conference on Artificial Intelligence for Good (AI4G)\u003c/em\u003e (pp. 55–62). IEEE. https://doi.org/10.1109/AI4G50087.2020.9319390\u003c/p\u003e\n\u003cp\u003eChesney, R., \u0026amp; Citron, D. (2019). Deep fakes: A looming challenge for privacy, democracy, and national security. \u003cem\u003eCalifornia Law Review\u003c/em\u003e, 107(6), 1753–1819. https://doi.org/10.15779/Z38R49J16H\u003c/p\u003e\n\u003cp\u003eCNN. (2024, February 4). Finance worker pays out $25 million after video call with deepfake “chief financial officer”. \u003cem\u003eCNN International\u003c/em\u003e. https://edition.cnn.com/2024/02/04/asia/deepfake-cfo-scam-hong-kong-intl-hnk/index.html\u003c/p\u003e\n\u003cp\u003eCONAR. (2020). \u003cem\u003eBrazilian Code of Advertising Self-Regulation\u003c/em\u003e. CONAR.\u003c/p\u003e\n\u003cp\u003eDataSciConnect. (2023). Generative AI deepfakes: Volkswagen ad revives Brazilian singer, raises concerns. \u003cem\u003eDataSciConnect\u003c/em\u003e. https://datasciconnect.com/generative-ai-deepfakes-volkswagen-ad-revives-brazilian-singer-raises\u003c/p\u003e\n\u003cp\u003eDixon, C. (2024). \u003cem\u003eRead and write your own\u003c/em\u003e. Random House.\u003c/p\u003e\n\u003cp\u003eEisenhardt, K. M. (1989). Building theories from case study research. \u003cem\u003eAcademy of Management Review\u003c/em\u003e, 14(4), 532–550. https://doi.org/10.2307/258557\u003c/p\u003e\n\u003cp\u003eEuropean Commission. (2021). \u003cem\u003eWhite paper on artificial intelligence: A European approach to excellence and trust\u003c/em\u003e. Publications Office of the European Union.\u003c/p\u003e\n\u003cp\u003eEuropean Commission. (2022). \u003cem\u003eThe Digital Services Act\u003c/em\u003e. https://commission.europa.eu/strategy-and-policy/priorities-2019-2024/europe-fit-digital-age/digital-services-act_en\u003c/p\u003e\n\u003cp\u003eEuropean Commission. (2024). \u003cem\u003eEU Artificial Intelligence Act\u003c/em\u003e. https://artificialintelligenceact.eu\u003c/p\u003e\n\u003cp\u003eFaden, R. R., Beauchamp, T. L., \u0026amp; Kass, N. E. (2014). \u003cem\u003eA history and theory of informed consent\u003c/em\u003e. Oxford University Press. https://doi.org/10.1093/med/9780190495679.001.0001\u003c/p\u003e\n\u003cp\u003eFolha de S.Paulo. (2023). Deepfake failed in a commercial with Elis Regina by ignoring her past. \u003cem\u003eFolha de S.Paulo\u003c/em\u003e. https://www1.folha.uol.com.br/colunas/manuela-cantuaria/2023/07/a-tecnologia-deepfake-falhou-no-comercial-com-elis-regina.shtml\u003c/p\u003e\n\u003cp\u003eFurze, L. (2024a, August 23). AI images are already more realistic than you think. \u003cem\u003eLeon Furze Blog\u003c/em\u003e. https://leonfurze.com/2024/08/23/ai-images-are-already-more-realistic-than-you-think/\u003c/p\u003e\n\u003cp\u003eFurze, L. (2024b). Deepfake game. \u003cem\u003eLeon Furze Blog\u003c/em\u003e. https://leonfurze.com/deepfake-game/\u003c/p\u003e\n\u003cp\u003eFuture of Life Institute. (2024). \u003cem\u003eThe act texts | EU Artificial Intelligence Act\u003c/em\u003e. https://artificialintelligenceact.eu/the-act/\u003c/p\u003e\n\u003cp\u003eGeorge, A. S., \u0026amp; George, A. S. H. (2023). Deepfakes: The evolution of hyper-realistic media manipulation. \u003cem\u003ePartners Universal Innovative Research Publication\u003c/em\u003e, 1(2), 48–54. https://doi.org/10.5281/zenodo.10148558\u003c/p\u003e\n\u003cp\u003eGoodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., iCourville, A., \u0026amp; Bengio, Y. (2014). Generative adversarial nets.In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, \u0026amp; K. Q. Weinberger (Eds.), \u003cem\u003eAdvances in neural information processing system\u003c/em\u003es \u003cem\u003e(Vol. 27)\u003c/em\u003e. http://arxiv.org/abs/1406.2661\u003c/p\u003e\n\u003cp\u003eGrewal, D., Guha, A., \u0026amp; Becker, M. (2024). AI is changing the world: For better or for worse? \u003cem\u003eJournal of Macromarketing\u003c/em\u003e, 44(4), 870–882. https://doi.org/10.1177/02761467241254450\u003c/p\u003e\n\u003cp\u003eGupta, G., Raja, K., Gupta, M., Jan, T., Whiteside, S. T., \u0026amp; Prasad, M. (2023). A comprehensive review of deepfake detection using advanced machine learning and fusion methods. \u003cem\u003eElectronics\u003c/em\u003e, 12(4), 865. https://doi.org/10.3390/electronics12040865\u003c/p\u003e\n\u003cp\u003eISO/IEC. (2023). \u003cem\u003eISO/IEC 23894:2023 – Information technology — Artificial intelligence —Guidance on risk management\u003c/em\u003e. International Organization for Standardization.\u003c/p\u003e\n\u003cp\u003eHam, C.-D. (2017). Exploring how consumers cope with online behavioral advertising. \u003cem\u003eInternational Journal of Advertising\u003c/em\u003e, 36(4), 632–658. https://doi.org/10.1080/02650487.2016.1239878\u003c/p\u003e\n\u003cp\u003eHarris, R. G., \u0026amp; Carman, J. M. (1983). Public regulation of marketing activity: Part I: Institutional typologies of market failure. \u003cem\u003eJournal of Macromarketing\u003c/em\u003e, 3(1), 49–58.\u003c/p\u003e\n\u003cp\u003eHasan, H. R., Salah, K., Jayaraman, R., Yaqoob, I., \u0026amp; Omar, M. (2024). NFTs to combat deepfakes and fake metaverse content. \u003cem\u003eInternet of Things\u003c/em\u003e, 25, 101133. https://doi.org/10.1016/j.iot.2024.101133\u003c/p\u003e\n\u003cp\u003eHeidari, A., Navimipour, N. J., Dag, H., \u0026amp; Unal, M. (2023). Deepfake detection using deep learning methods: A systematic and comprehensive review. \u003cem\u003eWIREs Data Mining and Knowledge Discovery\u003c/em\u003e, 13(6), e1528. https://doi.org/10.1002/widm.1528\u003c/p\u003e\n\u003cp\u003eHermann, E. (2022). Leveraging artificial intelligence in marketing for social good—An ethical perspective. \u003cem\u003eJournal of Business Ethics\u003c/em\u003e, 179(1), 43–61. https://doi.org/10.1007/s10551-021-04843-y\u003c/p\u003e\n\u003cp\u003eHutt, M. D., Mokwa, M. P., \u0026amp; Shapiro, S. J. (1986). The politics of marketing: Analyzing the parallel political marketplace. \u003cem\u003eJournal of Marketing\u003c/em\u003e, 50(1), 40–51.\u003c/p\u003e\n\u003cp\u003eHwang, Y., Kim, H., \u0026amp; Rho, J. (2021). Deepfakes and scientific knowledge dissemination. \u003cem\u003eScientific Reports\u003c/em\u003e, 11, 17769. https://doi.org/10.1038/s41598-021-96343-0\u003c/p\u003e\n\u003cp\u003eInternetLab. (2017). Black Mirror: “Be Right Back” (S02 E01), mourning and digital inheritance. \u003cem\u003eInternetLab\u003c/em\u003e. https://internetlab.org.br/pt/opiniao/black-mirror-be-right-back-s02-e01-luto-e-heranca-digital/\u003c/p\u003e\n\u003cp\u003eJing, T. W., \u0026amp; Murugesan, R. K. (2021). Protecting data privacy and preventing fake news and deepfakes in social media via blockchain technology. In \u003cem\u003eAdvances in Cyber Security: ACeS 2020\u003c/em\u003e (pp. 674–684). Springer. https://doi.org/10.1007/978-981-15-9129-7_48\u003c/p\u003e\n\u003cp\u003eKietzmann, J., Lee, L., McCarthy, I., \u0026amp; Kietzmann, T. (2020). Deepfakes: Trick or treat? \u003cem\u003eBusiness Horizons\u003c/em\u003e, 63(2), 135–146. https://doi.org/10.1016/j.bushor.2019.11.006\u003c/p\u003e\n\u003cp\u003eKietzmann, J., Mills, A., \u0026amp; Plangger, K. (2021). Deepfake: Perspectives on the future “reality” of advertising and branding. \u003cem\u003eInternational Journal of Advertising\u003c/em\u003e, 40(3), 473–485. https://doi.org/10.1080/02650487.2020.1834211\u003c/p\u003e\n\u003cp\u003eKshetri, N., Dwivedi, Y. K., Davenport, T. H., \u0026amp; Panteli, N. (2024). Generative artificial intelligence in marketing: Applications, opportunities, challenges, and research agenda. \u003cem\u003eInternational Journal of Information Management\u003c/em\u003e, 75, 102716. https://doi.org/10.1016/j.ijinfomgt.2023.102716\u003c/p\u003e\n\u003cp\u003eKumar, V., Kotler, P., Gupta, S., \u0026amp; Rajan, B. (2024). EXPRESS: Generative AI in marketing: Promises, perils, and public policy implications. \u003cem\u003eJournal of Public Policy \u0026amp; Marketing\u003c/em\u003e. https://doi.org/10.1177/07439156241286499\u003c/p\u003e\n\u003cp\u003eLacerda, I., \u0026amp; Mattos, T. R. (2020). Be Right Back: Humans, artificial intelligence, and Dasein in Black Mirror. \u003cem\u003eCommunications, technologies et développement\u003c/em\u003e. https://doi.org/10.4000/ctd.3353\u003c/p\u003e\n\u003cp\u003eLessig, L. (2006). \u003cem\u003eCode: And other laws of cyberspace (Version 2.0)\u003c/em\u003e. Basic Books.\u003c/p\u003e\n\u003cp\u003eLivedune. (2023). 70 years of Volkswagen in Brazil: Campaign features Elis Regina. \u003cem\u003eLivedune Blog\u003c/em\u003e. https://livedune.com/br/blog/70-anos-de-volkswagen-no-brasil-campanha-elis-regina/\u003c/p\u003e\n\u003cp\u003eMatos, J. de P., Gondo, M., Mota, L. S. E., \u0026amp; Horta, P. M. (2023). Food advertising aimed at children and adolescents in Brazil: Longitudinal analysis of complaints to CONAR. \u003cem\u003eCiência \u0026amp; Saúde Coletiva\u003c/em\u003e, 28(7), 1959–1970. https://doi.org/10.1590/1413-81232023287.14752022\u003c/p\u003e\n\u003cp\u003eMazis, M. B. (1997). Marketing and public policy: Prospects for the future. \u003cem\u003eJournal of Public Policy \u0026amp; Marketing\u003c/em\u003e, 16(1), 139–143. https://doi.org/10.1177/074391569701600112\u003c/p\u003e\n\u003cp\u003eMazis, M. B., Settle, R. B., \u0026amp; Leslie, D. C. (1977). Elimination of phosphate detergents and psychological reactance. \u003cem\u003eJournal of Marketing Research\u003c/em\u003e, 14(3), 390–395. https://doi.org/10.1177/002224377701400315\u003c/p\u003e\n\u003cp\u003eMeira, W. (2024). \u003cem\u003eLecture on Artificial Intelligence and Governance\u003c/em\u003e [Video]. YouTube. https://www.youtube.com/live/O9czIDgeVFI\u003c/p\u003e\n\u003cp\u003eMetropoles. (2023). Fake images of the Pope wearing a stylish coat go viral on social media. \u003cem\u003eMetrópoles\u003c/em\u003e. https://www.metropoles.com/mundo/religiao-int/imagens-falsas-do-papa-com-casaco-estiloso-viralizam-nas-redes-sociais\u003c/p\u003e\n\u003cp\u003eMigalhas. (2023). Conar files lawsuit against Volks for commercial featuring Elis Regina. \u003cem\u003eMigalhas\u003c/em\u003e. https://www.migalhas.com.br/quentes/392385/conar-arquiva-processo-contra-volks-por-comercial-com-elis-regina\u003c/p\u003e\n\u003cp\u003eMittelstadt, B. D., Allo, P., Taddeo, M., Wachter, S., \u0026amp; Floridi, L. (2016). The ethics of algorithms: Mapping the debate. \u003cem\u003eBig Data \u0026amp; Society\u003c/em\u003e, 3(2), 1–21. https://doi.org/10.1177/2053951716679679\u003c/p\u003e\n\u003cp\u003eMirsky, Y., \u0026amp; Lee, W. (2021). The creation and detection of deepfakes: A survey. \u003cem\u003eACM Computing Surveys\u003c/em\u003e, 54(1), 1–41. https://doi.org/10.1145/3392209\u003c/p\u003e\n\u003cp\u003eMotta, G. da S. (2022). What is a technological paper? \u003cem\u003eRevista de Administração Contemporânea\u003c/em\u003e, 26(Sup. 1), e220208. https://doi.org/10.1590/1982-7849rac2022220208.por\u003c/p\u003e\n\u003cp\u003eNaffi, N., Charest, M., Danis, S., Pique, L., Davidson, A., Brault, N., Bernard, M., \u0026amp; Barma, S. (2023). Empowering youth to combat malicious deepfakes and disinformation: An experiential and reflective learning experience informed by personal construct theory. \u003cem\u003eJournal of Constructivist Psychology\u003c/em\u003e, 1–22. https://doi.org/10.1080/10720537.2023.2294314\u003c/p\u003e\n\u003cp\u003eNightingale, S. J., \u0026amp; Wade, K. A. (2022). Identifying and minimizing the impact of fake visual media: Current and future directions. \u003cem\u003eMemory, Mind \u0026amp; Media\u003c/em\u003e, 1, e15. https://doi.org/10.1017/mem.2021.15\u003c/p\u003e\n\u003cp\u003eOpenAI. (2024). \u003cem\u003eSora\u003c/em\u003e. https://openai.com/sora\u003c/p\u003e\n\u003cp\u003ePittz, T. G., Steiner, S. D., \u0026amp; Pennington, J. R. (2020). An ethical marketing approach to wicked problems: Macromarketing for the common good. \u003cem\u003eJournal of Business Ethics\u003c/em\u003e, 164, 301–310. https://doi.org/10.1007/s10551-019-04277-7\u003c/p\u003e\n\u003cp\u003ePrajapati, S., Gadhari, J., Sawant, T., Kini, J., \u0026amp; Solanki, S. (2023). Strengthening supply chain integrity with blockchain-based anti-counterfeiting measures. In \u003cem\u003e2023 International Conference on Innovative Data Communication Technologies and Application (ICIDCA)\u003c/em\u003e (pp. 786–790). IEEE.\u003c/p\u003e\n\u003cp\u003eRana, M., Nobi, M. N., Murali, B., \u0026amp; Sung, A. H. (2022). Deepfake detection: A systematic literature review. \u003cem\u003eIEEE Access\u003c/em\u003e, 10, 25494–25513. https://doi.org/10.1109/ACCESS.2022.3154404\u003c/p\u003e\n\u003cp\u003eRicciardi Celsi, L. (2023). The dilemma of rapid AI advancements: Striking a balance between innovation and regulation by pursuing risk-aware value creation. \u003cem\u003eInformation, 14\u003c/em\u003e(12), 645. https://doi.org/10.3390/info14120645\u003c/p\u003e\n\u003cp\u003eRini, R. (2020). \u003cem\u003eDeepfakes and the epistemic backstop\u003c/em\u003e [Preprint]. PhilArchive. https://philarchive.org/rec/RINDAT-2\u003c/p\u003e\n\u003cp\u003eShahzad, H. F., Rustam, F., Flores, E. S., Vidal Mazón, J. L., de la Torre Díez, I., \u0026amp; Ashraf, I. (2022). A review of image processing techniques for deepfakes. \u003cem\u003eSensors, 22\u003c/em\u003e(12), 4556. https://doi.org/10.3390/s22124556\u003c/p\u003e\n\u003cp\u003eScherer, L. D., \u0026amp; Pennycook, G. (2020). Who is susceptible to online health misinformation? \u003cem\u003eAmerican Journal of Public Health\u003c/em\u003e, 110(S3), S276–S277. https://doi.org/10.2105/AJPH.2020.305908\u003c/p\u003e\n\u003cp\u003eSeneviratne, O. (2022). Blockchain for social good: Combating misinformation on the web with AI and blockchain. In \u003cem\u003eProceedings of the 14th ACM Web Science Conference 2022\u003c/em\u003e (pp. 435–442). ACM. https://doi.org/10.1145/3501247.3531549\u003c/p\u003e\n\u003cp\u003eShapiro, S. J. (2006). Macromarketing: Origins, development, current status, and possible future direction. \u003cem\u003eEuropean Business Review\u003c/em\u003e, 18(4), 307–321. https://doi.org/10.1108/09555340610677449\u003c/p\u003e\n\u003cp\u003eShoot Online. (2023, August 15). The best work you may never see: Volkswagen and AlmapBBDO reunite the mother. \u003cem\u003eShoot Online\u003c/em\u003e. https://www.shootonline.com/video/best-work-you-may-never-see-volkswagen-almapbbdo-reunite-mother-and-daughter-duet-ai\u003c/p\u003e\n\u003cp\u003eShoaib, M. R., Wang, Z., Ahvanooey, M. T., \u0026amp; Zhao, J. (2023). Deepfakes, misinformation, and disinformation in the era of frontier AI, generative AI, and large AI models. In \u003cem\u003e2023 International Conference on Computer and Applications (ICCA)\u003c/em\u003e (pp. 1–7). IEEE. https://doi.org/10.1109/ICCA58330.2023.10342321\u003c/p\u003e\n\u003cp\u003eSudhakar, K. N., \u0026amp; Shanthi, M. B. (2023). Deepfake: An endanger to cyber security. In \u003cem\u003e2023 International Conference on Sustainable Computing and Smart Systems (ICSCSS)\u003c/em\u003e (pp. 1542–1548). IEEE. https://doi.org/10.1109/ICSCSS56776.2023.10170989\u003c/p\u003e\n\u003cp\u003eSuperior Electoral Court. (2024). \u003cem\u003eTSE prohibits the use of artificial intelligence to create and spread false content in elections\u003c/em\u003e. https://www.tse.jus.br/comunicacao/noticias/2024/Fevereiro/tse-proibe-uso-de-inteligencia-artificial-para-criar-e-propagar-conteudos-falsos-nas-eleicoes\u003c/p\u003e\n\u003cp\u003eThe Drum. (2023, August 15). Heartwarming tribute or ethical breach? AI-powered Volkswagen ad sparks controversy. \u003cem\u003eThe Drum\u003c/em\u003e. https://www.thedrum.com/news/2023/08/15/heartwarming-tribute-or-ethical-breach-ai-powered-volkswagen-ad-sparks-controversy\u003c/p\u003e\n\u003cp\u003eTurkle, S. (1997). \u003cem\u003eLife on the screen: Identity in the age of the Internet\u003c/em\u003e. Simon \u0026amp; Schuster.\u003c/p\u003e\n\u003cp\u003eUOL. (2023, July 14). Elis, rights of the dead, and AI: Will Conar set limits for advertising? \u003cem\u003eUOL Economia — Mídia e Marketing\u003c/em\u003e. https://economia.uol.com.br/newsletters/midia-e-marketing/2023/07/14/julgamento-de-comercial-com-elis-pode-ser-historico-para-a-publicidade.htm\u003c/p\u003e\n\u003cp\u003eUOL News. (2024, September 14). Tabata Amaral takes legal action after photos are manipulated and published on social media. \u003cem\u003eUOL Notícias\u003c/em\u003e. https://noticias.uol.com.br/ultimas-noticias/agencia-estado/2024/09/14/tabata-amaral-aciona-a-justica-por-crime-de-injuria-eleitoral.htm\u003c/p\u003e\n\u003cp\u003eVainio-Pekka, H., Agbese, M., Jantunen, M., Vakkuri, V., Mikkonen, T., Rousi, R., \u0026amp; Abrahamsson, P. (2023). The role of explainable AI in the research field of AI ethics. \u003cem\u003eACM Transactions on Interactive Intelligent Systems\u003c/em\u003e, 13(4), Article 67. https://doi.org/10.1145/3599974\u003c/p\u003e\n\u003cp\u003eVerdoliva, L. (2020). Media forensics and deepfakes: An overview. \u003cem\u003eIEEE Journal of Selected Topics in Signal Processing\u003c/em\u003e, 14(5), 910–932. https://doi.org/10.1109/JSTSP.2020.2972392\u003c/p\u003e\n\u003cp\u003eWhittaker, L., Letheren, K., \u0026amp; Mulcahy, R. (2021). The rise of deepfakes: A conceptual framework and research agenda for marketing. \u003cem\u003eAustralasian Marketing Journal\u003c/em\u003e, 29(3), 204–214. https://doi.org/10.1177/1839334921999479\u003c/p\u003e\n\u003cp\u003eYin, R. K. (2014). \u003cem\u003eCase study research: Design and methods\u003c/em\u003e (5th ed.). Sage Publications.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Artificial Intelligence (AI), deepfake, Advertising and Propaganda, blockchain, digital ethics, digital marketing","lastPublishedDoi":"10.21203/rs.3.rs-8118450/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8118450/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe increasingly sophisticated quality of deepfakes makes it difficult for the public to distinguish between real and manipulated content, raising serious ethical and legal questions. This paper explores the ethical and legal complexities of using deepfake technology by analyzing the Volkswagen Brazil advertising case, which recreated the image of the late 80\u0026rsquo;s Brazilian popular music singer Elis Regina. Based on the decision of the Brazilian National Council for Advertising Self-Regulation (CONAR), the paper discusses the challenges advertisers face when integrating advanced technologies into marketing activities, while accounting for the perspectives of different stakeholders, including heirs of the artists, on consent for the use of posthumous images. The findings indicate a pressing need for more straightforward, comprehensive ethical and legal guidelines that address both explicit consent and the social and cultural impacts of AI use in advertising, especially for vulnerable audiences. The study underscores the importance of coordinated regulation adapted to the complexities of emerging technologies such as deepfakes, highlighting the nascent stage of current laws and regulations and the need for an approach that encompasses legal, social, and cultural aspects. As a potential technological solution, this study highlights the use of blockchain technology to address these issues.\u003c/p\u003e","manuscriptTitle":"Ethical and legal implications of using deepfakes in commercials: analysis of the Volkswagen Brazil case and the possibilities of blockchain","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-04 15:20:31","doi":"10.21203/rs.3.rs-8118450/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1c2feb21-04ba-4d4d-9a37-113dab35357e","owner":[],"postedDate":"December 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-03T11:18:12+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-04 15:20:31","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8118450","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8118450","identity":"rs-8118450","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}