Denmark has proposed to amend its Copyright Act in an effort to protect individuals from the risks posed by AI-generated deepfakes.[1] Unsurprisingly, this has made waves in the IP world, and the world in general. Over the last couple of months, several headlines have emerged stating that Denmark’s proposed amendments would ‘give people copyright to their own features’[2] or ‘let citizens copyright their own likeness’. These headlines certainly sound exciting (at least to an IP lawyer), but what do they really mean in practice? Will citizens need to take steps to ‘acquire the copyright’ in their likeness, or does it arise automatically? Is the person the protected ‘work’ as under traditional copyright principles? Would that mean that our likeness could become an intellectual property right, licensable to others (cue Black Mirror episode)?
As discussed in our previous article, deepfakes have been causing issues for celebrities and public figures for quite some time now. However, we are now increasingly seeing deepfakes used to target private individuals.[3] As deepfake technology continues to evolve, and its misuse becomes more prevalent, it is reassuring to see governments take action to establish legal protection for their citizens.
Denmark’s proposed copyright reform is considered further below, as well as the potential for similar reform in the UK.
Denmark’s proposed amendments
In short, the Danish government has proposed the following key amendments to the Danish Copyright Act to give individuals rights over their personal characteristics in AI-generated deepfakes[4]:
- Section 65 a. Realistic digitally generated imitations of a performing artist’s or performer’s artistic performance must not be made available to the public without the consent of the performing artist or performer.
- Section 73 a. Realistic digitally generated imitations of a natural person’s personal, physical characteristics may not be made available to the public without the consent of the person imitated.
These amendments would provide individuals with legal grounds to demand the removal of unauthorised deepfake content from media platforms.
There are, of course, exceptions to the above. For example, the rule under s73 a would not apply to imitations that are primarily expressions of caricature, parody, pastiche etc, however, these exceptions fall outside the scope of this article.
The overlap with copyright
Traditionally (and simplistically) copyright protects original ‘works’ and grants the author exclusive rights to reproduce, distribute, perform, display, and adapt those works.
Denmark’s proposed amendments would introduce a sort of hybrid right that sits between personality rights and copyright. Although the right to demand removal of deepfake content would be embedded in the Danish Copyright Act, it would not be dependent on the creation of a ‘work’ in the traditional sense. Instead, it would automatically protect individuals against unauthorised imitation or digital manipulation of their personal features, such as voice, appearance or movement.
What about the UK?
As a general principle, the UK does not recognise ‘image rights’.[5] As discussed in our previous article, individuals subject to a ‘deepfake attack’ in the UK must rely on a patchwork of causes of action, including passing off and criminal and civil law protections. Additionally, the UK GDPR protects against the processing of any information that can identify a living person directly or indirectly, including biometric data. It gives individuals the right to object to their data being used, however, the enforcement of the GDPR is often regarded as feeble. Denmark’s proposal offers an innovative legal solution, however, adopting similar principles in the UK would require either a new legal framework, or a significant expansion of existing rights.
The UK could undergo legislative reform to introduce rights which mirror Denmark’s proposed sections 65a and 73a, giving individuals automatic protection over their likeness, voice and characteristics. This could be by way of a statutory digital personality right, introduced either via a new act or incorporated into the Copyright, Designs and Patents Act (CDPA) 1988. The CDPA already provides performers with certain rights to authorise the making of recordings of their performance.[6] However, these rights are limited to live performances and do not extend to AI-generated content. Denmark’s proposal could fill this gap by recognising digital impersonation as a form of infringement. However, any legislative reform in response to deepfakes would require careful consideration to avoid unintended consequences.
For example, it is notable that Denmark’s proposed amendment requires that consent to publicly disclose a deepfake be given by the person imitated. This is likely intended to curtail collective licencing mechanisms and to ensure that control remains with the individual. However, this requirement for individual consent could still potentially be circumvented as the individual’s ‘control’ could be contractually overridden. By way of illustration, the person imitated could be contractually bound to always provide consent to the use of their personal features. Accordingly, if the UK adopted a similar approach, individuals would need to exercise caution when negotiating contracts involving the use of their likeness, for example, ambassador or endorsement agreements.
Alternatively, existing UK IP rights could be expanded to address deepfakes. For example, under the law of passing off, the threshold for establishing goodwill could potentially be lowered in cases involving personal image misuse. Under trade mark law, perhaps an individual could register their personal features as a non-traditional trade mark – although would that mean that new trade mark applications are needed as one progressively ages? Evidently, these suggestions would open all sorts of floodgates and serve to highlight the complexity of adapting existing legal frameworks to emerging technologies.
The good news is that the UK is questioning whether its current legal framework provides adequate protection in the context of artificial intelligence. Notably, the UK government recently ran a consultation on copyright and artificial intelligence which closed on 25 February 2025.[7] Although the primary focus of the consultation is on using copyright works to train AI models, the consultation sought views on proposals to introduce personality rights in the UK.[8] It will be interesting to see whether Denmark’s proposal features in the consultation outcome and whether there is consensus that a new right is needed to protect individuals from digital impersonation.
Conclusion
Denmark’s proposal to protect individuals from deepfakes by embedding rights within its Copyright Act is a thought-provoking response to a growing digital threat. The UK’s current IP framework does not adequately safeguard an individual’s likeness from exploitation by AI and perhaps Denmark’s approach will prompt reform. As AI continues to challenge legal boundaries, the need for clear protections over personal identity has never been more prevalent – though how such protections will integrate with existing UK law remains to be seen.
References
[1] <https://www.gov.uk/government/consultations/copyright-and-artificial-intelligence> accessed 16 September 2025.
[2] <https://www.gov.uk/government/consultations/copyright-and-artificial-intelligence/copyright-and-artificial-intelligence> Paragraphs 175 and 178, accessed 16 September 2025.
[3] Fenty v Arcadia Group Brands Ltd [2015] EWCA Civ 3 [29] (Kitchin LJ).
[4] Copyright, Designs and Patents Act 1988, s182.
[5] <https://kum.dk/aktuelt/nyheder/bred-aftale-om-deepfakes-giver-alle-ret-til-egen-krop-og-egen-stemme> accessed 6 October 2025.
[6] Miranda Bryant, ‘Denmark to tackle deepfakes by giving people copyright to their own features’ (The Guardian, 27 June 2025) <https://www.theguardian.com/technology/2025/jun/27/deepfakes-denmark-copyright-law-artificial-intelligence> accessed 6 October 2025.
[7] Resemble.AI, ‘Q1 2025 Deepfake Incident Report: Mapping Deepfake Incidents’ (2025) <https://www.resemble.ai/wp-content/uploads/2025/04/ResembleAI-Q1-Deepfake-Threats.pdf> accessed 16 September 2025.
[8] English translation (machine-translated) of the Proposal –https://www.ft.dk/samling/20241/almdel/kuu/bilag/232/3050901.pdf
