Learning from Hollywood? Why the US right of publicity can only serve as a partial model for Europe’s deepfake regulation

The Deepfake Problematic

The phenomenon of deepfakes challenges existing image and voice protection. Unlike before, AI systems can now generate audio and video content that mimics a real person to the point of being indistinguishable from the original. This creates problems across several areas: politicians are mimicked, women and children are victims of so-called deepfake pornography and, in the entertainment industry, singers’ voices and actors’ images are used without consent to create new, fake content.

Jurisdictions address these challenges differently. In most continental European countries, such as France and Germany, image and voice rights are regarded as personal rights rather than intellectual property rights. Denmark and Cyprus, by contrast, have respectively proposed and introduced a new neighbouring right within copyright law protecting image and voice, an approach that the Netherlands are also discussing but might not adopt.

In the US, meanwhile, the right of publicity has already existed for decades and is treated as an intellectual property right, even though, in practice, it is almost exclusively invoked by public figures whose identity has a demonstrable market value. Given the long running debate in Europe on whether or not certain personality rights like image and voice rights should be modelled on the US publicity right, this blog post argues that the US right of publicity can only serve as a partial model for Europe’s deepfake regulation.

Image and Voice Rights in Europe

At European level, Article 8 of the ECHR forms the central legal basis by enshrining the right to privacy, a right which, following the Hannover ruling, now also encompasses the right to one’s own image. In addition, the provisions of the GDPR apply in certain circumstances (Romero Moreno, 2024).

In some Member States, like Germany and France, the protection goes even further: there, the right to one’s own image is also recognised as partly patrimonial, meaning it can, to a certain extent, be commercially exploited (Brüggemeier, 2022).

It is noteworthy that no distinction whatsoever is made here between public figures and private individuals. Rather, the underlying assumption is that anyone entitled to a commercial right to their image or voice must also possess a corresponding market value. Such a market value cannot be readily assumed in the case of ‘ordinary’ people, but it can be assumed in the case of public figures. Rather than concluding that public figures forfeit their right to privacy, however, European courts instead weigh the personality rights of public figures against competing rights, such as freedom of expression, freedom of the press and artistic freedom, within a proportionality assessment.

Still, in its European form, this right is mostly conceived solely as a right of defence, not as a right that can be actively exercised. Consequently, any potential licensing or transfer of these rights as market value falls outside their scope.

National approaches in the EU

To close gaps in protection, Denmark has proposed new legislation. This new sui generis right has been structured as a neighbouring right within copyright law and covers both performing artists and ‘ordinary’ people.

However, it remains controversial whether this is in fact an independent intellectual property right with active licensing possibilities, or whether its structure is, at its core, still characterised by personality rights (Karttunen, 2026). Accordingly, the draft bill appears primarily to codify and clarify previously unwritten principles underpinning the general right of personality. Although its inclusion in copyright legislation may initially seem dubious, that legislative choice implicitly acknowledges the particular risks to which performing artists are exposed.

Cyprus has already modelled its approach on the Danish model, albeit in a less sophisticated form. As in Denmark, it remains unclear whether the new right can be licensed or transferred (Synodinou, 2026).

A similar regulation is currently under discussion in the Netherlands. P. Bernt Hugenholtz, however, takes a critical view of this proposal, primarily on the grounds that classifying the issue as a form of intellectual property law does not adequately reflect the actual risks involved. Accordingly, copyright law is simply the wrong starting point for regulating deepfakes; the focus should instead be on media regulation, as well as data protection and personality rights. This view is also shared by the Dutch Commissie Auteursrecht, which likewise concluded that an intellectual property right is unsuitable for combating the misuse of voice and image in contexts such as pornography, fraud and deception.

The US Publicity Right

In some US states, the right of publicity is already regarded as an intellectual property right. It emerged as a counterpart to the right of privacy and, in this conception, protects public figures like celebrities and politicians.

According to US case law, public figures, by virtue of their high-profile status, are deemed to have partially waived their privacy rights and are therefore deprived of their protection. To close the resulting gap in protection, the right of publicity was created as a separate statutory right grounded in intellectual property law.

However, there is yet no uniform, nationwide recognition of this right, resulting in a disparate and largely incoherent regulatory landscape. California, for example, is the frontrunner in this regard and, not least due to its proximity to Hollywood, offers the strongest protection of publicity rights.

It should also be emphasised here that the publicity right, as such, does not recognise any moral rights. Consequently, it lacks a non-transferable core and a component relating to personality rights (Dogan and Lemley, 2006). The absence of moral rights is best explained by the US’s long-standing resistance to the Berne Convention and its eventual accession in 1988 without specific legislative recognition of moral rights, with Congress instead relying on a ‘legal patchwork’ to comply with Article 6bis of Berne.

Assessment and Outlook

In my view, Denmark’s approach is a first step in the right direction, at least as far as the protection of performers is concerned. A separate intellectual property right with possible licensing options seems appropriate here, as the market value of artists’ voices and likenesses is growing steadily. It is now a reality that both a person’s voice and their likeness can be utilized and marketed as independent economic assets.

It remains questionable, however, whether ‘ordinary’ individuals should also be covered by such a right. In some respects, such individuals may already be adequately protected by the general right of personality and privacy. Here, the decisive criterion should be the actual market value associated with a person’s identity, as well as the question of whether an intellectual property right relating to that identity would have such value, irrespective of whether the person in question is a public figure or not. A two-tiered protection should be envisaged: On the one hand, the right to one’s own image and voice, with its connection to moral rights, would remain in place; on the other hand, a separate intellectual property right could be established, enabling authors and performers to commercially exploit their persona without encroaching on the core area of moral rights.

In the light of these considerations, however, one must still agree with Hugenholtz that intellectual property rights, as such, should not remain the sole means of addressing the issue. Deepfake pornography and deepfakes in a political context fall under different regulatory frameworks, which, as such, should not be enshrined in intellectual property law. Deepfake pornography, which according to studies accounts for almost 96 per cent of all deepfake cases, constitutes a serious violation of the privacy rights of the individuals concerned, whether they are public figures or not. The mere ability to freely dispose of one’s own image or voice is, in principle, not sufficient here; rather, protection under criminal law and platform regulation will also be required, at the very least (Kira, 2026). In the case of political deepfakes, however, the situation is different: it is not primarily the person depicted who requires protection, but instead the public itself, namely from the fact that what is shown online does not actually correspond to reality.