Advances in artificial intelligence, particularly in the generation of synthetic voices using voice cloning, are raising increasing legal questions in practice. Technically, it is now possible to generate digital voices that sound deceptively similar to the original voice of a natural person, such as prominent voice actors. From a legal perspective, the question arises as to whether and to what extent the reproduction of a human voice constitutes an infringement of the actor’s personality right or other legal positions. Case law is beginning to address such issues from a practical perspective. A ruling by the Regional Court of Berlin dated 20 August 2025 (file number: 2 O 202/24) clearly affirms legal protection for the human voice in the context of synthetic reproductions and at the same time comments on the calculation of damages.

Personality rights as laid down in the German Constitution (Art. 2 (1), Art. 1 (1) GG) protects the individual identity and integrity of a person. The voice of a person is also regarded as a means of expression of personality worthy of protection in the established case law of the Federal Court of Justice (e.g. judgement of 18 June 1996 – file number: VI ZR 543/94 – ‘Marlene Dietrich’). This applies even if the voice is only imitated or reproduced. The prerequisite is its recognisability, meaning that the connection to the natural person is obvious.

Personality rights & voice acting 

In the context of synthetic voices generated by AI tools, an infringement already occurs if the artificially generated voice evokes a clear association with the natural person without their consent. From a legal perspective, it is irrelevant whether the sound was generated synthetically. The decisive factors for assessing whether there has been a violation of personality rights are the effect on the audience and the type of use. Particularly relevant here are situations in which the artificially generated voice is disseminated publicly, for example in the context of advertising or political expression. In these cases, the interests of the person concerned, i.e. the owner of the original voice, will regularly prevail and a legal violation will be affirmed.

Copyright & voice acting

In addition, the use of voices in AI may raise copyright issues. Voice actors are generally not classified as authors, but as performing artists by ancillary copyright within the meaning of the Copyright Act (“Leistungsschutzrechtsinhaber” – § 73 UrhG). The voice acting may therefore be legally protected independently of the actual sound recording. If such voice recordings are used without consent, this may constitute an infringement of ancillary copyright. Although the scope of protection of ancillary copyright is narrower than that of copyright-protected works, this may also be a possible starting point for legal action. 

Data Protection Law & Voice acting 

Finally, Data Protection Law can be considered a protective medium for dubbing. The voice is classified as so-called biometric data and enjoys the protection of special personal data (Art. 9 GDPR). If voice recordings are used entirely without the consent of the persons concerned and there is no contractually justified reason for their use, this may constitute unauthorised use of personal data. 

No free availability of voices

In its ruling of 20 August 2025, the Regional Court of Berlin (file number: 2 O 202/24) issued a landmark decision in the area of AI use of voices of well-known voice actors. A well-known dubbing artist took action against the unauthorised use of his voice. The actor considered his rights to have been infringed by the use of AI software to generate a deceptively similar synthetic voice and use it in two publicly accessible videos – including a political commentary and an advertising video for an online shop.

The YouTuber who was sued argued in court that he had not specifically instructed the use of the claiming voice actor but was merely looking for an ‘authentic voice with a heroic sound’. However, this objection did not impress the judges. The court affirmed a violation of the voice actor’s personal rights and, at the same time, an economically relevant infringement. The reasoning behind the ruling states that a person’s voice is not only a part of their personality that is worthy of protection, but also economically exploitable and licensable. The imitation leads to an unwanted identification of the speaker with the content of the videos, even if the voice was digitally generated in this specific case.

The YouTuber had already issued a cease-and-desist declaration out of court. In addition, the voice actor was awarded damages in the form of a licence fee of EUR 4,000 (EUR 2,000 per video). The calculation was based on the usual remuneration for voice acting in a comparable context. The court emphasised that the amount should be based on the individual market position and exclusivity of the imitated voice.

Practical consequences of the judgement

The decision of the Regional Court of Berlin makes it clear that a technical generation of a voice by AI falls within the scope of personality rights. The decisive factor is not the method of voice production, but the actual effect on third parties. For companies, platform operators and developers of synthetic speech models, this results in even greater review and documentation obligations. In contractual practice, the question of how the use of the voice is regulated in contracts with dubbing artists will also arise more frequently.

The court decision is also significant because every person’s voice – not just that of voice actors – is legally protected. In addition, the court’s considerations can be applied to other aspects of a person’s personality (such as physical appearance, gestures and movements). Finally, the question arises how to deal with the personality rights of deceased personalities. In this case, the post-mortem personality right (“postmortales Persönlichkeitsrecht”) may come into play when imitating the voice/appearance. However, it needs to be clarified as to how long after death claims exist and who among the circle of heirs/relatives/close persons can take action against such infringements. 

Conclusion: One’s own voice is not only a means of expression, but also an economically relevant object of protection. The ruling of the Regional Court of Berlin shows that even synthetic, AI-generated voice imitations can violate personality rights and are subject to legal restrictions.