Denmark and AI Deepfakes: Protecting Voice, Likeness and Digital Identity
Viral posts say Denmark has granted citizens “copyright over their face, voice and body.” The reality is more nuanced: the country has proposed a copyright-law amendment intended to make it easier to stop the unauthorised public sharing of realistic AI-generated imitations of people.
What has Denmark actually proposed?
Generative artificial intelligence can now create convincing videos, images and audio recordings of people who never said the words attributed to them and never took part in the events shown. This material, commonly described as deepfakes, can be harmless or creative. It can also be used for fraud, reputational attacks, public manipulation and the unauthorised imitation of a person’s voice or appearance.
That risk has fuelled the viral claim that Denmark is “the first country in Europe to give people copyright in their own face, voice and body.” The phrase is memorable, but not legally exact. The Danish initiative is not conventional copyright in a person’s body or facial features as creative works. It is a proposal to use the Copyright Act to create a new copyright-style protection against realistic, digitally generated imitations of a person.
A key qualification is necessary at the time of publication. Available information indicates that the measure remains a draft bill rather than unquestionably enacted and operative law. Denmark notified the draft through the EU TRIS system, and the European Commission has commented on it. Any reference to a “new Danish law” should therefore be understood as a description of a legislative initiative and its potential direction, not as a definitive account of established court or administrative practice.
Who would be protected, and against what?
The proposed new section 73a concerns realistic, digitally generated imitations of a natural person’s personal physical characteristics. In practical terms, this includes appearance and voice. The intended legal effect is that such an imitation may not be made available to the public without the consent of the person being imitated. Based on the available explanations, the proposal is designed to protect all natural persons, including non-Danish nationals.
The proposal also adds a separate layer for performers and artists. Proposed section 65a would protect them against the public availability of realistic, digitally generated imitations of their performances or artistic presentations without consent. This matters particularly to actors, musicians, presenters, comedians, dubbing artists and other professionals whose voices, likenesses or performances have commercial value.
The model addresses a practical weakness in traditional remedies. A victim of a deepfake may have claims based on data protection, privacy, reputation, unfair competition or criminal law, but it is not always easy to identify one clear rule that directly covers a digital imitation of appearance or voice. Denmark’s proposal therefore seeks to offer a recognisable legal basis for challenging and removing the material.
What could an affected person seek?
The key point is that the proposed prohibition targets the public availability of a realistic digital imitation without consent. It is not a blanket ban on every use of AI, nor does it automatically punish everyone who creates a satirical image. The focus is on unauthorised publication or sharing with the public.
Where a person’s voice or likeness is used in a deceptive deepfake posted on a social-media platform, the proposed rules would provide a clearer foundation to:
- request removal of the disputed material;
- report the content to a platform as illegal;
- demand that unauthorised use of the digital imitation cease;
- pursue other available remedies, including damages where the relevant legal requirements are met.
The proposal aims to establish a clearer legal basis that an imitated person could use to seek removal of realistic digital imitations from storage and sharing platforms, including social media.
The phrase “heavy fines” used in viral posts should nevertheless be treated with caution. Available information supports a framework against unauthorised publication and for reporting content to platforms, while the final rules on procedure, liability and sanctions depend on the adopted statutory text and its application.
Satire, parody and the right to criticise
Any deepfake regime needs to balance two legitimate interests: protection for the person whose voice or appearance is imitated, and freedom of expression. The Danish draft therefore provides exceptions for material that is mainly caricature, satire, parody, pastiche, criticism of power or social criticism.
The exception is not unlimited. It should not protect an imitation that amounts to misinformation capable of causing concrete, serious harm to the rights or significant interests of others. A clearly satirical clip is therefore different from a fabricated audio message in which a chief executive tells an employee to make an urgent payment to a new bank account.
This line also matters for political speech. Deepfake satire may serve legitimate public debate, while synthetic content presented as authentic and capable of seriously misleading voters, consumers or business partners requires a different legal response.
The relationship with the AI Act and platforms
Denmark’s proposal must be viewed alongside the wider EU framework. The EU AI Act provides transparency obligations for certain AI-generated or AI-manipulated content, including deepfakes. The European Commission is also working on a code of practice concerning the marking and labelling of AI-generated content. The Digital Services Act, in turn, establishes mechanisms for notifying online platforms about illegal content.
Protection against deepfakes will therefore not rest on a single rule. Depending on the facts, data-protection law, privacy, consumer protection, criminal law, copyright and related rights, as well as contracts between businesses, agencies and performers, may all be relevant. Denmark’s approach matters because it attempts to supplement this fragmented landscape with a clearer right against unauthorised digital imitation.
What businesses should do now?
Businesses should not wait for the final Danish law before putting controls in place for AI avatars, cloned voices and synthetic videos. The most important measure is obtaining clear, documented consent whenever a real person’s recognisable voice, face, likeness, recording or performance is used.
Sound contractual and compliance practice includes:
- specific consent identifying the person, purpose, channels, territory and period of use;
- clear labelling of AI-generated or AI-altered content where required;
- contractual warranties from agencies, production companies and AI vendors that they hold the necessary rights;
- a rapid internal process for receiving, assessing and removing potentially unlawful deepfake content;
- staff training to recognise voice-cloning fraud, especially in finance, procurement and payment workflows.
Conclusion
Denmark has not literally granted citizens conventional copyright in their face, voice or body. Its proposal does, however, represent an important effort to protect digital identity through a right clear enough to help people challenge harmful, realistic AI-generated imitations.
The central lesson for individuals, creators and businesses is that deepfakes are no longer only a technology or reputation issue. They sit at the intersection of personality rights, intellectual property, data protection, platform governance and AI compliance. Consent, transparent labelling and an effective takedown process are becoming fundamental elements of responsible generative-AI use.
Follow for more legal insights:
