Authored By: Maria Fejzulla
Salford University
The rapid development of artificial intelligence has fundamentally transformed the modern digital landscape. Among the most controversial products of this technological evolution is the emergence of “deepfakes,” a form of synthetic media generated through artificial intelligence that manipulates audio, video, or images to create highly realistic but fabricated content.[1] While artificial intelligence has undoubtedly produced significant social and economic benefits, deepfake technology has simultaneously introduced unprecedented threats to privacy, dignity, democratic integrity, and personal autonomy. Deepfakes possess the capacity to distort reality itself, allowing individuals to be falsely represented in compromising, defamatory, or misleading situations without their knowledge or consent. The consequences of such manipulation extend beyond mere reputational harm and increasingly implicate fundamental human rights protected under domestic constitutions and international legal instruments.
The growing prevalence of deepfake technology has exposed serious deficiencies within existing legal systems. Traditional laws concerning defamation, privacy, obscenity, harassment, and cybercrime were not designed to address the unique challenges posed by artificial intelligence-generated identity manipulation.[2] The speed, realism, and global accessibility of deepfakes render conventional legal remedies inadequate, particularly where victims suffer irreversible emotional, social, and professional harm before judicial intervention becomes possible. Moreover, the cross-border nature of digital platforms complicates enforcement and raises difficult questions concerning jurisdiction, accountability, and freedom of expression. This article argues that deepfake technology poses a substantial threat to fundamental human rights, including privacy, dignity, equality, and psychological integrity, and that existing legal frameworks remain insufficient to address these harms effectively. It further contends that a comprehensive human rights-based regulatory framework is necessary to balance technological innovation with the protection of individual freedoms and democratic values.
Deepfakes are typically created using machine learning systems known as Generative Adversarial Networks (“GANs”), which analyse large quantities of data to replicate facial expressions, voices, and gestures with remarkable accuracy.[3] Initially developed for entertainment and research purposes, deepfake technology has rapidly evolved into a tool capable of producing convincing synthetic content that is difficult for ordinary viewers to distinguish from authentic material. The accessibility of such technology has dramatically increased in recent years due to the widespread availability of open-source software and mobile applications, enabling even non-experts to generate manipulated media.[4] As a result, deepfakes have proliferated across social media platforms, political campaigns, pornography websites, and digital communication networks.
One of the most alarming uses of deepfakes concerns non-consensual sexually explicit content, particularly targeting women. Studies consistently indicate that the overwhelming majority of online deepfake pornography involves female victims whose identities are digitally inserted into explicit videos without consent.[5] Such conduct constitutes a severe violation of human dignity, bodily autonomy, and privacy. The psychological consequences for victims often include anxiety, humiliation, reputational damage, social isolation, and professional harm.[6] In many cases, victims experience forms of digital sexual violence comparable to offline abuse. Deepfake pornography therefore raises significant concerns under international human rights law, particularly Article 12 of the Universal Declaration of Human Rights (“UDHR”), which protects individuals against arbitrary interference with privacy, family, and reputation.[7] Similar protections are provided under Article 17 of the International Covenant on Civil and Political Rights (“ICCPR”).[8]
The human rights implications of deepfakes extend beyond individual privacy violations and increasingly threaten democratic governance and public trust. Deepfakes have the potential to manipulate elections, incite political violence, and disseminate misinformation on an unprecedented scale.[9] Fabricated videos depicting political leaders engaging in criminal, inflammatory, or unethical conduct may rapidly circulate online before verification mechanisms can respond effectively. Such manipulation undermines the public’s ability to distinguish truth from fabrication and weakens confidence in democratic institutions.[10] The spread of false information through deepfake technology also implicates the collective right of societies to access reliable information and participate meaningfully in democratic processes.
Furthermore, deepfake technology disproportionately affects vulnerable groups and may reinforce structural discrimination. Women, journalists, activists, and public figures are particularly susceptible to targeted harassment through manipulated media.[11] Female politicians and journalists frequently encounter sexually explicit deepfakes intended to intimidate, silence, or discredit them in public life. This phenomenon raises concerns under international equality and non-discrimination principles, including those recognised in the Convention on the Elimination of All Forms of Discrimination Against Women (“CEDAW”).[12] The gendered nature of deepfake abuse demonstrates that artificial intelligence technologies can reproduce and amplify existing social inequalities if left unregulated.
Despite the severity of these harms, existing legal frameworks remain fragmented and inadequate. Traditional defamation law offers limited protection because victims often struggle to establish damages before manipulated content spreads globally across digital platforms.[13] Defamation remedies also tend to focus on reputational harm rather than violations of dignity, autonomy, or psychological integrity. Privacy laws similarly encounter limitations because many jurisdictions lack clear statutory recognition of digital identity rights or consent-based protections against synthetic media manipulation.[14] Moreover, legal proceedings are often lengthy and ineffective against content capable of reaching millions of viewers within hours.
In the context of non-consensual deepfake pornography, some jurisdictions have attempted legislative responses. Several states in the United States have enacted laws criminalising sexually explicit deepfakes or prohibiting their dissemination without consent.[15] However, these laws vary considerably in scope and enforcement mechanisms, producing inconsistent protection for victims. Additionally, many statutes require proof of malicious intent or demonstrable harm, thereby creating evidentiary burdens that may discourage victims from pursuing legal remedies.[16] The absence of uniform federal legislation in the United States further complicates accountability and enforcement.
The European Union has adopted a comparatively proactive approach through the Artificial Intelligence Act (“AI Act”), which introduces transparency obligations for certain AI-generated content, including deepfakes.[17] Under the AI Act, providers must disclose when content has been artificially generated or manipulated, thereby promoting transparency and reducing deceptive uses of synthetic media.[18] While the legislation represents an important regulatory development, critics argue that transparency requirements alone may not sufficiently address malicious deepfake abuse, particularly where harmful content spreads rapidly through anonymous accounts or foreign platforms.[19]
In India, deepfakes are currently addressed through a combination of provisions under the Information Technology Act 2000, the Indian Penal Code, and intermediary guidelines issued under the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules 2021.[20] However, India lacks dedicated legislation specifically targeting deepfake technology. Existing provisions concerning impersonation, obscenity, defamation, and cybercrime may apply indirectly, yet they fail to comprehensively address the distinct challenges associated with AI-generated synthetic media.[21] The absence of explicit regulation creates uncertainty regarding platform liability, victim remedies, and evidentiary standards.
One of the most difficult legal questions surrounding deepfake regulation concerns the relationship between synthetic media and freedom of expression. Deepfakes are not inherently harmful; they may serve legitimate purposes in entertainment, satire, education, artistic expression, and accessibility technologies.[22] Excessively broad regulation may therefore risk suppressing lawful speech and creativity. International human rights law recognises freedom of expression as a fundamental right under Article 19 of the ICCPR and Article 19 of the UDHR.[23] Consequently, legal frameworks must carefully balance the prevention of harmful deepfakes against the protection of legitimate expressive freedoms.
However, freedom of expression is not absolute. International law permits restrictions where necessary to protect the rights and reputations of others, national security, public order, or public health.[24] Deepfakes that involve fraud, harassment, non-consensual pornography, electoral manipulation, or incitement to violence clearly fall outside the scope of protected democratic discourse. The challenge therefore lies not in determining whether regulation is permissible, but rather in designing proportionate mechanisms capable of distinguishing harmful synthetic media from legitimate expression.
A human rights-based regulatory framework offers the most effective approach to addressing these challenges. Such a framework should prioritise the protection of dignity, privacy, equality, and democratic integrity while preserving lawful innovation and freedom of speech. First, states should enact legislation specifically criminalising malicious deepfakes involving sexual exploitation, impersonation, electoral deception, and fraudulent conduct.[25] These laws should adopt consent-based standards rather than requiring proof of reputational or financial damage. The absence of consent itself should constitute a central element of liability in cases involving manipulated identity abuse.
Second, digital platforms should bear greater responsibility for detecting and removing harmful deepfake content. Social media companies currently play a central role in the dissemination of manipulated media yet frequently respond inconsistently to reports of abuse.[26] Mandatory due diligence obligations should require platforms to implement effective detection systems, rapid complaint mechanisms, and transparent moderation procedures. Additionally, victims should possess accessible rights to seek immediate removal of harmful content without excessive procedural burdens.
Third, governments and international organisations should promote technological safeguards capable of identifying synthetic media. Watermarking systems, digital provenance standards, and authenticity verification tools may help distinguish genuine content from manipulated material.[27] While technical solutions alone cannot eliminate deepfake abuse, they may reduce the speed and scale of misinformation dissemination. International cooperation is particularly necessary because digital content frequently crosses national borders beyond the jurisdictional reach of domestic regulators.
The role of international human rights institutions is also increasingly significant. The United Nations and regional human rights bodies should develop guidelines addressing artificial intelligence and synthetic media within existing human rights frameworks. Such standards would assist states in harmonising regulatory approaches while ensuring compliance with international obligations concerning privacy, equality, dignity, and freedom of expression.[28] Given the global nature of digital platforms, fragmented national responses are unlikely to provide effective long-term protection.
Moreover, legal responses must recognise that deepfake abuse is not solely a technological issue but fundamentally a human rights issue. The manipulation of an individual’s face, voice, and identity without consent strikes at the core of personal autonomy and dignity. Unlike traditional misinformation, deepfakes possess an unparalleled capacity to create emotional and psychological harm because they simulate authenticity with extraordinary realism. Victims may suffer profound social consequences even where fabricated content is eventually disproven.[29] The law must therefore evolve beyond conventional frameworks that treat such harm merely as reputational or economic injury.
In conclusion the rapid advancement of artificial intelligence ensures that deepfakes will become increasingly sophisticated and difficult to detect in the coming years. Without timely legal intervention, the risks posed by synthetic media will continue to expand across political, social, and personal domains. Existing laws remain insufficiently equipped to address the scale, speed, and complexity of AI-generated identity abuse. Deepfakes threaten fundamental human rights, including privacy, dignity, equality, and democratic participation, while simultaneously eroding public trust in authentic information. Although freedom of expression and technological innovation must remain protected, these values cannot justify the exploitation, harassment, and manipulation enabled by malicious synthetic media. A comprehensive human rights-based regulatory framework is therefore essential to ensure accountability, protect victims, and preserve democratic integrity in the digital age. Ultimately, the challenge posed by deepfakes is not merely technological but civilisational: societies must determine whether artificial intelligence will be governed in service of human dignity or permitted to undermine the very concept of truth itself.
Table of Legislation
India Information Technology Act 2000
EU Artificial Intelligence Act 2024
UN Convention on the Elimination of All Forms of Discrimination against Women (Adopted 18 December 1979, entered into force 3 September 1981) 1249 UNTS 13
UN International Covenant on Civil and Political Rights (Adopted 16 December 1966, entered into force 23 March 1976) 999 UNTS 171
UN Universal Declaration of Human Rights (Adopted 10 December 1948 UNGA Res 217 A(III))
Bibliography
Books
Ajder H, Patrini G, Cavalli F and Cullen L, The State of Deepfakes: Landscape, Threats, and Impact (Deeptrace Labs 2019)
Murray A, Information Technology Law (4th edn, Oxford University Press 2019)
Schick N, Deepfakes and the Infocalypse (Monoray 2020)
Journal Articles
Citron DK and Chesney R, ‘Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security’ (2019) 107 California Law Review 1753
Delfino R, ‘Pornographic Deepfakes: The Case for Federal Criminalization’ (2021) 13 Drexel Law Review 887
Edwards L, ‘Regulating Deepfakes’ (2023) European Human Rights Law Review 45
Goodfellow I and others, ‘Generative Adversarial Nets’ in Z Ghahramani and others (eds), Advances in Neural Information Processing Systems (MIT Press 2014)
McGlynn C, Rackley E and Houghton R, ‘Image-Based Sexual Abuse’ (2017) 37 Oxford Journal of Legal Studies 225
Sources and Online Articles
Chesney R and Citron D, ‘Deep Fakes and the New Disinformation War’ Foreign Affairs (2019)
Gupta A and Sawhney RS, ‘Deepfakes and the Indian Legal Landscape’ (Internet Freedom Foundation, 2023)
Organisation for Economic Co-operation and Development, OECD AI Principles (2019)
United Nations Educational, Scientific and Cultural Organisation, Recommendation on the Ethics of Artificial Intelligence (2021)
[1] Danielle Keats Citron & Robert Chesney, “Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security,” 107 Cal. L. Rev. 1753 (2019).
[2] ibid
[3] Ian Goodfellow et al., “Generative Adversarial Nets,” Advances in Neural Information Processing Systems (2014).
[4] Henry Ajder et al., The State of Deepfakes: Landscape, Threats, and Impact (Deeptrace Labs 2019).
[5] ibid
[6] Clare McGlynn et al., “Image-Based Sexual Abuse,” 37 Oxford J. Legal Stud. 225 (2017).
[7] United Nations, Universal Declaration of Human Rights, art. 12 (1948).
[8] United Nations, International Covenant on Civil and Political Rights, art. 17 (1966).
[9] Bobby Chesney & Danielle Citron, “Deep Fakes and the New Disinformation War,” Foreign Affairs (2019).
[10] ibid
[11] Nina Schick, Deepfakes and the Infocalypse (Monoray 2020).
[12] Convention on the Elimination of All Forms of Discrimination Against Women, United Nations (1979).
[13] Citron & Chesney, supra note 1.
[14] Andrew Murray, Information Technology Law (4th ed. 2019).
[15] Rebecca Delfino, “Pornographic Deepfakes: The Case for Federal Criminalization,” 13 Drexel L. Rev. 887 (2021).
[16] ibid
[17] European Union Artificial Intelligence Act, 2024.
[18] ibid
[19] Lilian Edwards, “Regulating Deepfakes,” 2023 Eur. Hum. Rts. L. Rev. 45.
[20] Information Technology Act, No. 21 of 2000, India.
[21] Apar Gupta & Ria Singh Sawhney, “Deepfakes and the Indian Legal Landscape,” Internet Freedom Foundation (2023).
[22] Robert Chesney & Danielle Citron, “Disinformation on Steroids,” 95 Foreign Aff. 147 (2019).
[23] UDHR art. 19; ICCPR art. 19.
[24] ICCPR art. 19(3).
[25] Delfino, supra note 15.
[26] Schick, supra note 11.
[27] Organisation for Economic Co-operation and Development, OECD AI Principles (2019).
[28] United Nations Educational, Scientific and Cultural Organisation, Recommendation on the Ethics of Artificial Intelligence (2021).
[29] McGlynn et al., supra note 6.
