Regulating DEEPFAKES: Can COPYRIGHT LAW keep up with AI Manipulation?

When seeing is no longer believing and AI blurs the line between creation and imitation, can copyright law survive the rise of deepfakes?

Imagine seeing a video of someone saying or doing something they never actually did, but it looks so real that you almost believe it. That’s the power of deepfakes.

These are AI‑generated images, audio, or videos created by training machines on real footage, making it possible to fabricate events that never happened.

As the technology behind deepfakes becomes increasingly advanced and accessible, the risks are growing rapidly. Deepfakes are now used to spread false information, manipulate public opinion, and invade personal privacy by putting people in situations they never consented to. The impact can be serious. Fake political videos can influence elections and divide communities. Non‑consensual deepfake pornography can ruin reputations and relationships. The flood of misinformation can even erode trust in social media, news outlets, and even everyday conversations.

All of this raises a significant question: “Can copyright law, originally built to protect traditional creative works like books, music, and films, keep up with this new wave of AI manipulation? Or is it simply not enough to tackle the unique challenges deepfakes present?”

Understanding Deepfakes

What Are Deepfakes?

Deepfakes are a type of synthetic media, like images, videos, or audio recordings, that are generated or manipulated using advanced artificial intelligence to look and sound real. The word itself is a blend of “deep learning” and “fake,” which perfectly captures how these creations work: AI systems study large amounts of real footage and then use that knowledge to produce new, fabricated content that closely imitates real people and events.

At their core, deepfakes rely on deep learning models, especially a type of neural network known as Generative Adversarial Networks (GANs). A GAN functions like a creative competition between two AI systems:

• The generator that creates fake data, and

• The discriminator that spots fake data

Over time, after thousands of training cycles, the generator becomes so skilled that its outputs can fool even sophisticated detectors, not just the human eye.

In practical terms, deepfakes can be created through several techniques, including:

• Face-swapping: Replacing one person’s face with another’s in a video or photo.

• Voice cloning: Recreating a person’s voice using a short audio sample, often just a few seconds long.

• Full-body or avatar synthesis: Generating realistic movements, facial expressions, and gestures to create entire videos of people doing things they never did.

What makes deepfakes especially powerful and dangerous is how convincingly these tools blend visual and audio data. They can capture not only how someone looks, but how they speak, react, and behave. Without specialized detection tools, many deepfakes are nearly impossible for the average person to identify, blurring the line between genuine human activity and AI-generated illusion.

Types of Deepfake Content

Deepfakes come in many forms. They are not a single, uniform type of media; instead, they exist along a wide spectrum, from harmless creativity to serious harm. Understanding these categories helps clarify why regulating deepfakes is so complex.

1. Entertainment and Satire: Not all deepfakes are malicious. Some are created purely for fun, artistic experimentation, or parody. In the entertainment world, creators often use deepfake tools to insert familiar faces into movie scenes, recreate historical figures, or produce light-hearted sketches. A popular example is the web series Sassy Justice, which uses deepfake technology to satirize well-known political figures. In these contexts, the intention is humor or commentary, although they still raise questions about consent and transparency.

2. Political Misinformation: One of the more alarming uses of deepfakes is the fabrication of political content. AI-generated videos can make public figures appear to endorse policies they never supported, deliver speeches they never gave, or behave in ways that could damage their credibility. These deepfakes can spread rapidly across social media, influencing public opinion, misleading voters, and potentially destabilizing democratic processes. Even when debunked, the initial shock can have lasting effects on public trust.

3. Non-Consensual Explicit Content: This is widely regarded as the most harmful form of deepfake content. AI tools are often used to create sexually explicit videos of individuals, mostly women and minors, without their knowledge or consent. Victims frequently face social stigma, reputational damage, emotional distress, and long-term psychological harm. Entire documentaries and investigative reports have exposed how easily a person’s face can be taken from a simple photo and placed into explicit materials. This category has fueled global calls for stronger legal protections.

4. Commercial Use; Advertising, Endorsements, and Impersonation: Deepfakes are increasingly appearing in the commercial landscape. Brands, marketers, and content creators sometimes use AI-generated likenesses of celebrities or influencers to promote products, occasionally without obtaining permission. In other cases, malicious actors create fake endorsement videos to scam consumers. These scenarios blur ethical boundaries and have even resulted in legal disputes over unauthorized use of personal likeness, false advertising, and consumer deception.

Why Deepfakes Are Unique

Deepfakes are not just another form of media manipulation; they represent a fundamentally new challenge for technology, society, and the law. Several factors make them distinct from traditional forms of altered content:

1. Speed and Accessibility of Creation: Creating realistic media used to require specialized skills, expensive software, and hours of painstaking work. Today, AI-powered tools have democratized the process. With just a personal computer or smartphone, anyone with an internet connection can produce deepfake content within minutes. This rapid creation accelerates the production and dissemination of synthetic media, making harmful content easier to produce and more difficult to control.

2. Increasing Difficulty of Detection: Early deepfakes often had noticeable flaws, unnatural facial movements, mismatched lighting, or distorted audio. Modern AI, however, can mimic tiny details such as lip synchronization, subtle facial micro-expressions, and unique voice patterns, producing content that is nearly indistinguishable from reality. This makes it challenging for ordinary viewers, and sometimes even advanced detection algorithms, to recognize fakes, allowing deceptive content to circulate widely before it can be verified or removed.

3. Blurring the Line Between Reality and Fabrication: Unlike traditional editing tools, deepfakes can generate entirely new events that never occurred. They don’t simply alter reality; they invent it. This capability undermines trust in digital media, including news reporting, social media posts, and shared personal videos. Over time, this erosion of trust can have broader societal consequences, from skepticism about legitimate information to difficulty establishing truth in legal or journalistic contexts.

4. Scalability and Virality: Deepfakes can be mass-produced and distributed across social media platforms at unprecedented speed. A single video can be replicated, remixed, and shared millions of times, amplifying its influence far beyond traditional manipulated media. This scalability transforms what might have once been a localized hoax into a potentially global phenomenon.

5. Multifaceted Threats: Deepfakes impact multiple domains simultaneously, including personal privacy, political integrity, financial security, and social cohesion. Unlike older forms of media manipulation that were often isolated to one context, deepfakes can be weaponized across personal, commercial, and political spheres, creating overlapping risks that are harder to address through conventional legal frameworks.

We’ve seen that Deepfakes are more than just technological novelties; they raise profound legal and ethical questions. By blending visual and audio elements to create content that never actually occurred, deepfakes challenge our traditional understanding of authorship, ownership, and creative control. Unlike conventional media, which can usually be traced to a human creator, deepfakes often emerge from complex AI processes where the “author” may be a machine or a team orchestrating algorithms.

This raises a critical question: How does the law, particularly copyright law, respond when creativity is increasingly generated or mediated by artificial intelligence? Copyright law was historically designed to protect human innovation and reward authors for their original works, assuming a clear chain of authorship and ownership. But deepfakes blur those lines, forcing us to re-examine foundational assumptions about originality, fixation, and authorship, the very pillars upon which copyright stands.

It is in this context that we must examine the current copyright framework to understand where it aligns, where it falls short, and the challenges posed by AI-generated content.

Current Copyright Framework

Copyright law has long served as a cornerstone of intellectual property protection. Its primary purpose is to encourage creativity by granting authors exclusive rights over their original works, rewarding their labor and innovation. Traditionally, this framework assumed that creative works originate from identifiable human authors, whose efforts could be traced and protected. The rise of deepfakes challenges many of these assumptions.

When AI can generate highly realistic images, videos, or voices, the conventional ideas of authorship, originality, and ownership are thrown into question. Who owns a video created by an algorithm trained on thousands of existing clips? Can a machine be considered an author? These are questions that traditional copyright law struggles to answer.

Fundamentals of Copyright Protection

At the core of any copyright regime common in most jurisdictions is the doctrine that protects original works of human authorship fixed in a tangible medium of expression. That means copyright grants exclusive economic and moral rights to creators or right-holders over their works, covering reproduction, distribution, adaptation, public performance, and other derivative uses. Historically, copyright law emerged to incentivize human creativity: to reward authors for their labour, investment, and creative vision by granting them a monopoly over their works. In the pre-digital era, this model worked reasonably well: the chain of authorship, fixing, and rights allocation was relatively uncontroversial. This orthodox understanding requires:

• Originality: The work must demonstrate minimal creativity originating from a human mind.

• Fixation: The work must be captured in a tangible or sufficiently stable medium, such as a recording, manuscript, or digital file.

• Human Authorship: Copyright traditionally protects works created by people, not machines or automated processes.

For centuries, this model functioned adequately because expressive works could be traced to identifiable human creators. But generative AI offends this logic. Deepfake technology, powered by machine-learning architectures such as GANs (Generative Adversarial Networks) or diffusion models, produces content that simulates human creativity without any human-generated expression in the resulting output. Copyright law was not designed for a world where the “creator” may effectively be a machine optimising statistical patterns in data. Even the U.S. Copyright Office has affirmed repeatedly that non-human authorship cannot receive copyright protection, a stance reaffirmed in its 2023–2025 policy statements.

As we explore copyright in relation to deepfakes, it becomes clear that while the law provides some tools to address misuse, it was never intended to regulate content created or manipulated by artificial intelligence, a gap that is becoming increasingly evident.

Ownership & Rights

Once a work meets the requirements for copyright protection, the author, or the copyright owner if rights have been transferred, gains a bundle of exclusive rights. These rights are designed to give creators control over how their work is used, distributed, and adapted, while also providing legal remedies against unauthorized use.

The key rights typically include:

• Reproduction: This gives the copyright owner the exclusive ability to make copies of their work. Whether it’s a video, audio recording, image, or written content, no one else can reproduce it without the owner’s permission.

• Distribution: Copyright holders control how their work is shared with the public. This includes deciding whether the work can be sold, rented, streamed, or otherwise distributed. It allows authors to monetize their creations and manage the timing, scope, and method of public access.

• Public Performance: Authors can control when and how their work is performed publicly. This right covers activities such as broadcasts, live performances, or streaming of videos or music, ensuring creators are compensated for public use of their work.

• Adaptation & Derivative Works: Copyright gives owners the power to authorize or create adaptations of their work. This can include translations, remixes, adaptations into other formats (like turning a book into a film), or any derivative work that builds on the original.

• Public Display: Creators have the right to exhibit their work publicly. This applies to art in galleries, photographs in publications, or videos and graphics posted online. It ensures authors maintain control over the visual representation and public exposure of their creations.

Together, these rights form a protective framework that allows creators to manage and benefit from their work, while also giving them legal recourse if someone uses their creations without permission. In essence, copyright not only incentivizes creativity but also empowers authors to safeguard the integrity and commercial value of their works.

Exceptions & Limitations

While copyright grants creators significant control over their works, it is not absolute. Legal systems around the world recognize that the public interest sometimes outweighs exclusive rights, and therefore include exceptions and limitations that allow certain uses of copyrighted material without the owner’s permission. These provisions aim to strike a balance between protecting creators and promoting access to knowledge, cultural expression, and innovation.

• Fair Use (United States)

In the United States, the doctrine of fair use permits limited use of copyrighted works for purposes such as criticism, commentary, news reporting, teaching, scholarship, or research. Courts evaluate fair use by considering four key factors:

1. Purpose and character of the use – including whether the use is commercial or nonprofit, and whether it is transformative (i.e., adds new meaning or value rather than simply copying).

2. Nature of the copyrighted work – creative works receive stronger protection than purely factual works.

3. Amount and substantiality of the portion used – using small or less significant parts of a work may weigh in favor of fair use, while copying the “heart” of the work may not.

4. Effect on the market – if the use could negatively impact the work’s commercial value, it may weigh against fair use.

Fair use is highly flexible and context-dependent, allowing courts to consider the broader social and cultural impact of the use.

• Fair Dealing (Commonwealth and International Contexts)

In many Commonwealth countries, as well as in several other jurisdictions, a concept similar to fair use, known as fair dealing, governs exceptions to copyright. Fair dealing is generally narrower than U.S. fair use, often specifying particular purposes for which copyrighted works can be used without permission, such as:

• Research or private study

• Reporting current events

• Criticism, review, or commentary

• Parody and satire

Because fair dealing is more prescriptive, whether a use qualifies under this doctrine is typically more clear-cut, but it also offers less flexibility than the U.S. system.

• Parody and Satires

Many copyright laws explicitly recognize that parody and satire serve important social, cultural, and political functions. A parody or satirical work may reproduce elements of the original work to make a point, comment, or critique. However, whether a deepfake qualifies as a parody depends on multiple factors, including the creator’s intent, the context in which it is shared, and its impact on the original work. Courts often weigh these elements carefully when deciding whether a parody or satire falls within legal exceptions.

These exceptions and limitations illustrate that copyright is not meant to create a total monopoly over creative works. Instead, it seeks to balance creators’ rights with public interest, ensuring that copyrighted material can still be used in ways that support education, commentary, innovation, and cultural dialogue. In the context of deepfakes, these doctrines raise complex questions: when does a deepfake constitute permissible parody or commentary, and when does it cross the line into infringement?

Where Deepfakes Collide with Copyright Law

Deepfakes do more than raise ethical or social concerns; they expose significant legal gaps. Because deepfakes are often created by training AI systems on existing photographs, videos, music, or voice recordings, they inevitably intersect with copyright law, which regulates how creative works may be used, altered, and owned.

Unlike traditional creative processes, a deepfake creation often involves multiple layers of input: the original copyrighted material used for training, the AI model that processes the data, and the human user who initiates or guides the generation of the content. This complex chain makes it difficult to determine where copyright responsibility begins and ends. As a result, deepfakes sit uncomfortably within legal frameworks that were built for far simpler forms of authorship and creativity.

Below are some of the key areas where this collision becomes most apparent.

1. Issue of Authorship: Who Owns a Deepfake?

One of the most pressing legal questions surrounding deepfakes is authorship. Copyright law in most jurisdictions is built on the assumption that a work is the product of human creativity. As a result, copyright protection is typically reserved for works created by natural persons who exercise creative judgment and control over the final output. In the context of deepfakes, this assumption fails. Many deepfakes are generated largely, or in some cases entirely, by artificial intelligence systems. While a human may provide prompts, select inputs, or choose parameters, the final output is often produced through automated processes that rely on statistical pattern recognition rather than direct human expression.

Courts and regulatory bodies have begun to grapple with this issue. In the United States, for example, the Copyright Office has consistently maintained that works generated solely by AI, without sufficient human creative input, are not eligible for copyright protection. U.S. courts have echoed this position, reinforcing the idea that copyright cannot vest in a non-human author.

This leads to two questions:

• If an AI system generates a deepfake, who qualifies as the author? Is it the programmer who built the model, the company that trained it, or the user who supplied the prompt?

• If no human author can be identified, does the deepfake exist outside copyright protection altogether?

When no clear authorship exists, the deepfake may fall into a legal grey zone where no one holds enforceable copyright rights over the final output. This creates serious practical consequences. If harmful deepfake content lacks a clear owner, victims may struggle to rely on copyright law to demand takedowns or seek remedies, even when the content causes reputational, emotional, or economic harm.

2. Use of Real Persons’ Likeness:

A significant number of deepfakes involve the use of real people’s faces, voices, or identities. This is where the limits of copyright law become particularly clear. Copyright protects creative expression such as photographs, films, sound recordings, and other fixed works, but it does not protect a person’s face, voice, or identity in and of itself.

In practical terms, this means that copying or recreating someone’s likeness is not automatically a copyright violation unless the deepfake also reproduces a copyrighted work. For example, using a specific photograph, video clip, or audio recording without permission may infringe copyright, but merely replicating how someone looks or sounds often falls outside copyright’s scope.

This distinction is crucial in many deepfake cases. Non-consensual deepfake videos may cause severe reputational, emotional, and economic harm, yet copyright law may offer limited relief if the underlying material is not itself protected. Instead, these harms are more appropriately addressed under personality rights or rights of publicity, which protect an individual’s commercial and personal interest in their identity.

In jurisdictions such as the United States and parts of Europe, the right of publicity allows individuals, especially public figures, to control the commercial use of their name, image, likeness, or voice. For instance, using a celebrity’s face or voice in a deepfake advertisement without consent may violate their publicity rights, even if no copyrighted work is directly copied.

Conversely, if a deepfake is created using a copyrighted photograph or video that you own, you may pursue a copyright infringement claim alongside any personality-rights violation.

Recognizing the growing threat posed by deepfakes, some jurisdictions are beginning to explore new legal frameworks that grant individuals explicit rights over their digital likeness. These emerging approaches aim to extend protection beyond traditional copyright, reflecting the reality that faces and voices can now be replicated as easily as copyrighted works.

3. Copyright Infringement Claims

One of the clearest points at which deepfakes intersect with copyright law is through the unauthorized use of existing copyrighted works. While copyright may not protect a person’s identity, it does protect the creative works that are often used as raw material in deepfake creation.

If an AI-generated deepfake incorporates or is trained on copyrighted content, such as:

• A film clip owned by a studio or individual,

• A song or sound recording protected by copyright, or

• A photograph taken by a professional photographer,

The creator or distributor of that deepfake may be liable for copyright infringement. This is especially true where the deepfake reproduces, adapts, or creates derivative works from copyrighted material without the permission of the rights holder.

In many real-world cases, copyright infringement claims have become one of the most effective legal tools for addressing harmful deepfakes. Even where the underlying harm relates more to identity misuse, deception, or reputational damage, copyright law is often used strategically to secure takedowns, injunctions, or damages. This is because copyright frameworks are more established and easier to enforce than newer or less developed identity-based rights.

However, this approach also reveals a limitation: copyright law is frequently being used as a proxy to address harms it was never designed to regulate.

Liability Challenges: Who Is Responsible?

Beyond infringement itself, deepfakes raise difficult questions about legal responsibility. Unlike traditional copyright violations, deepfakes often involve multiple actors across different stages of creation and distribution.

The Creator: The individual who intentionally generates a deepfake is the most obvious source of liability. If the deepfake uses copyrighted material without authorization, the creator may be directly liable for infringement or for creating an unauthorized derivative work.

Platforms and Intermediaries: Social media platforms, websites, and hosting services may face secondary or contributory liability if they knowingly facilitate the distribution of infringing deepfakes. While many platforms rely on safe-harbor protections, these protections often depend on prompt action once notified of infringement.

AI System Developers: Whether developers of AI systems can be held responsible for deepfake misuse remains legally unsettled. In many jurisdictions, developers argue that they merely provide neutral tools and should not be liable for how users deploy them. Courts have yet to establish consistent standards for assessing responsibility at this level.

The issue becomes even more complex when the original creator cannot be identified, such as in cases involving anonymous users or cross-border distribution. In these situations, victims often turn to platforms or intermediaries as the most practical route for enforcement, using intellectual property laws to request takedowns and limit further dissemination.

The Gaps in Current Copyright Law

Despite the broad reach of copyright law, structural and doctrinal gaps emerge when that law confronts AI-driven deepfakes.

1. Copyright Does Not Protect Your Face or Voice

One of the most glaring omissions in traditional copyright law is that one’s likeness, a person’s face, voice, or other biometric or identity features, is not generally protected under copyright. Copyright protects expression (e.g., a photograph, a recording), not the underlying identity.

Thus, if a generative AI produces a synthetic video or audio that replicates a real person’s face or voice without using a copyrighted “source work,” existing copyright law may offer no redress. That gap leaves individuals, including public figures or private persons, vulnerable to misuse of their likeness via deepfakes.

This is precisely the problem identified by the U.S. Copyright Office (USCO) in its recent AI-and-copyright reports: the Office recognizes that “digital replicas” (i.e., deepfakes) can inflict serious harms, but that copyright law, as currently structured, does not adequately protect individuals’ identities. In other words, you do not “own” your face or voice under copyright, so deepfakes that misuse those elements fall through the cracks.

2. It is built for Traditional Media, not AI Manipulation

Copyright law was developed in an era of photographs, books, films, and physical media, where ownership, authorship, copying, adaptation, and distribution had relatively fixed, human-controlled pathways. The frameworks assume a human author, a work that is “fixed,” and a chain of ownership, custody, or licensing whenever that work is reused or adapted.

Generative AI disrupts this model: AI can ingest thousands of existing works (some copyrighted, many possibly unlicensed), learn stylized patterns, and output entirely new—but eerily realistic text, images, videos, or audio. These AI outputs may resemble, but not copy, the original works.

Moreover, the output may depict real persons in entirely fictional (and potentially harmful) settings, not by copying a pre-existing image or audio-visual recording, but by synthesizing a likeness from learned data. Traditional copyright lacks a doctrine for “synthetic identity reuse.” As many commentators have opined, copyright law is simply not “fit for purpose” for deepfake–driven identity manipulation.

3. Enforcement Challenges

Even where copyright could apply, for example, when deepfakes reuse copyrighted footage, audio, or images, enforcing it is fraught with challenges.

Authorship ambiguity: If AI generated the deepfake with minimal human intervention (e.g., a prompt), many jurisdictions may refuse to grant copyright, meaning no owner to sue. The 2025 decision by a U.S. appeals court rejecting a copyright claim over a purely AI-generated artwork is a recent example.

Attribution difficulties: Generative AI often trains on large datasets scraped from the internet. Tracing which input works gave rise to a particular output can be practically impossible. A recent forensic analysis study illustrated how complex and technical it is to attribute GAN-generated images to particular training datasets, a problematic prerequisite for establishing infringement or liability.

Jurisdictional fragmentation & international spread: Deepfakes can be created and distributed globally. Copyright law is territorial; even when there is infringement, cross-border enforcement is slow, costly, and legally complex.

Cost and burden of proof: Victims must identify the infringing content, track the infringer(s), and prove copying or unauthorized derivation, all of which can be resource-intensive.

Taken together, these gaps mean that many harmful deepfakes cannot realistically be challenged under existing copyright regimes, or if they can, the process may be so onerous as to render redress nominal.

What Other Laws Might Help?
Beyond Copyright: Other Legal Tools

Given these deficiencies, courts, regulators, and legislators have begun looking beyond copyright to alternative or complementary legal regimes.

1. Privacy Rights, or Right of Publicity & Personality Rights

A major avenue is the doctrine of right of publicity or personality rights. The right of an individual to control the commercial use of their name, image, voice, or likeness. In contexts such as synthetic endorsements, deepfake-based marketing, or commercial impersonation, these rights may offer stronger, more direct protection than copyright.

In many U.S. states (and in some other countries), right of publicity is available under statute or common law. It allows an individual to prevent unauthorized commercial uses of their persona, even if the deepfake does not copy any copyrighted work. However, a major limitation remains: the lack of a unified federal standard. Enforcement depends on the jurisdiction, and for non-commercial deepfakes (e.g., political satire, parody, misinformation), some of these rights may not apply, or may be subject to First-Amendment or free speech defenses.

2. Defamation, Privacy, and Harassment Laws

When deepfakes depict individuals saying or doing things they never did — especially in defamatory, harassing, or intimate contexts, defamation, privacy torts, and harassment laws can offer recourse.

For example, a deepfake portraying a private individual in an explicit context without consent might give rise to claims under non-consensual pornography laws, privacy laws, or harassment statutes (depending on the jurisdiction). Indeed, the harmful potential of deepfakes is not limited to identity theft or copyright infringement; false-speech deepfakes can destroy reputations, endanger personal safety, or facilitate blackmail, among other harms, which the copyright doctrine was never designed to address.

3. Emerging AI-Specific Regulations & Digital Replica Laws

Recognizing the inadequacy of existing frameworks, regulatory bodies are beginning to propose laws specific to AI. The U.S. Copyright Office itself, traditionally a purely copyright institution, recommended in its July 2024 “Digital Replicas” report that Congress enact a new federal right protecting individuals against the unauthorized distribution of “digital replicas” (i.e., deepfakes), irrespective of commerciality. Under the proposed law:

• It would apply to all individuals (not just celebrities),

• unauthorized distribution of deepfake likenesses (voice, face, image) would be prohibited, even non-commercial distribution,

• online platforms would have takedown obligations (mirroring copyright safe-harbor regimes),

• statutory damages, injunctive relief, and possibly criminal or civil penalties would be available.

Some jurisdictions may follow suit. Indeed, a recent European example is found in developments in Denmark: as of 2025, Denmark is pursuing proposed legislation to grant individuals explicit copyright-style control over their own face, voice, and body, to combat AI-generated deepfakes, a bold attempt to treat identity as intellectual property.

4. Data Protection and Digital Identity Frameworks

Beyond personality or intellectual property law, data and identity protection regimes can also play a role. Synthetic media often derive from massive datasets of personal data (faces, voices, biometric data). As courts and regulators grapple with identity theft, deception, and misuse, data protection laws or digital identity frameworks may become relevant, especially in jurisdictions with robust privacy or data protection regimes. Moreover, some technologists and scholars are proposing hybrid governance frameworks combining technical, legal, and policy safeguards.

For example, a recent proposal, the Digital Identity Rights Framework (DIRF), outlines a structured governance model for managing biometric and personality-based digital likeness attributes, integrating both legal and technical controls (consent, traceability, monetization, enforcement) to respond to generative AI identity risks. Such frameworks, if adopted, could offer a systemic, scalable alternative to relying solely on patchwork litigation under copyright, publicity, or privacy law.

So… Can Copyright Keep Up?

The short answer is yes, but only to a point.

Copyright law provides useful tools for addressing some of the harms caused by deepfakes, particularly where copyrighted images, videos, or audio are used without permission. It offers established enforcement mechanisms, such as takedown procedures and infringement claims, which have proven effective in limiting the spread of certain AI-generated content.

However, copyright was never designed to regulate synthetic media that can replicate human faces, voices, and identities with minimal or no human authorship. Many of the more serious harms caused by deepfakes, identity misuse, deception, reputational damage, and non-consensual exploitation fall outside copyright’s core purpose. As a result, copyright often functions as a workaround rather than a comprehensive solution.

Deepfakes expose fundamental gaps in existing legal frameworks. Questions around authorship, ownership, liability, and consent remain unresolved, especially when AI systems generate content autonomously or across borders. Relying on copyright alone risks stretching it beyond its intended scope and leaving victims without adequate protection.

Addressing deepfakes effectively will require a hybrid legal approach. Copyright law must be complemented by AI-specific regulations, stronger personality and publicity rights, data protection rules, and platform accountability measures. Together, these tools can create a more coherent system, one that protects creativity without sacrificing personal dignity, democratic integrity, or public trust.

In an era where seeing is no longer believing, the law must evolve just as rapidly as the technology it seeks to govern. Copyright can play a role, but it cannot stand alone.

---

About Legal Bytes

We are Adune Legal’s weekly Newsletter, which simplifies the Law for Busy Executives, Entrepreneurs, and Tech Enthusiasts interested in the legal aspects of Business, Technology, and Intellectual Property.

We love emails from our readers— reply to this email and let us know your thoughts and suggestions.

​WAIT!!!

Become a paid subscriber and access;

• Q&A sessions with Nneoma Grace via chats on Substack.

• Detailed Legal Templates and examples to save you time and legal fees

• Expert Interviews and Case Studies

Don’t miss out on these perks - subscribe today and start enjoying it!

Thanks for reading Legal Bytes​

​Adune Legal’s Team

P.S. Like Legal Bytes? Please forward us to a friend.

P.P.S. Was this publication forwarded to you? Sign up here & see previous publications.