In recent years, the rise of “deepfakes” has sparked major concern in international forums on intellectual property, image rights, and digital content legislation. This AI-based technology allows for the superimposition of human faces, voices, and expressions with increasing realism, creating fake but convincing videos or audio. The legal, ethical, and societal implications of this practice are the subject of intense debate.

What are deepfakes and why are they a problem?

The term “deepfake” combines “deep learning” and “fake” and refers to audiovisual content manipulated using machine learning algorithms to digitally recreate human faces or voices. While this technology has potential uses in entertainment, education, and even art, it has also been employed for malicious purposes: spreading misinformation, extortion, identity theft, and the creation of non-consensual sexual content.

Current legal challenges

One of the key issues is that existing legal frameworks in many countries are not equipped to handle deepfakes. Classical copyright law protects original works fixed in tangible form but not necessarily a person’s image or voice. Image rights legislation is fragmented and varies significantly across jurisdictions. Furthermore, prosecuting digital falsification crimes often requires proving intent to harm, which is difficult in anonymous online environments.

U.S. Copyright Office Report (2024)

In December 2024, the U.S. Copyright Office published a report stating that current laws are inadequate to address the risks posed by deepfakes. Key findings include:

• No comprehensive protection exists for image and voice as standalone rights.
• Manipulated works cannot be registered as original if they lack significant human involvement.
• The Office recommends Congress enact specific federal legislation addressing deepfakes.

Main risks associated with deepfakes

1. Violation of personal identity: The unauthorized use of someone’s face or voice can infringe on their right to self-image, especially if portrayed in offensive or deceptive contexts.
2. Impact on copyright: Manipulating a protected work (e.g., a film or song) to insert fictional elements may alter its integrity and create confusion about authorship.
3. Difficulty in traceability and accountability: Deepfake content is easily distributed anonymously via social media, making it difficult to identify the perpetrator.
4. Disinformation and social manipulation: Deepfakes can be used to create fake statements by public figures, affecting elections or triggering social unrest.

Regulatory proposals under discussion

1. Recognition of digital identity rights: Several countries are considering legal reforms to recognize image and voice as personal assets that can be protected independently of a work.
2. Mandatory labeling of synthetic content: Proposals are being made to require all AI-generated or manipulated content to carry a visible mark indicating its artificial origin.
3. Creation of digital provenance registries (blockchain): To guarantee authenticity, technologies for traceability and digital certification are being explored.
4. Platform liability: There is debate about whether platforms like YouTube, TikTok, or X should bear stricter obligations to detect and remove malicious deepfakes.
5. Specific criminal offenses: Some U.S. states have passed laws criminalizing the creation of deepfakes for defamatory or non-consensual sexual purposes.

Notable examples

• In 2023, a fake video showing Ukraine’s president announcing surrender to Russia went viral. Though quickly debunked, it temporarily caused public panic.
• Celebrities like Scarlett Johansson and Tom Hanks have publicly denounced the unauthorized use of their likeness in AI-generated advertisements.

Ethical considerations

• Informed consent: The use of a person’s image should require their explicit authorization, especially in sensitive contexts.
• Freedom of expression vs. rights protection: Regulation must strike a balance between creative freedom and safeguarding against manipulation.
• Digital education and media literacy: Public awareness and training are key to help people recognize and critically assess manipulated content.

Conclusion: Toward smart and preventive legislation

The deepfake phenomenon is one of the most pressing challenges to copyright and identity protection in the digital age. A multidisciplinary approach combining technological, legal, and educational measures is urgently needed. Creating a specific and harmonized legal framework at both national and international levels is essential to safeguard individual rights and preserve trust in today’s digital ecosystem.