Berlin Regional Court clarifies that the imitation of a celebrity voice generated by artificial intelligence can also infringe personal rights. In the case of German voice actor Manfred Lehmann, the use of his AI-generated voice on YouTube was deemed unlawful and the defendant was ordered to cease and desist and the plaintiff was awarded a licence fee in accordance with the principles of unjust enrichment law.
Background: When the AI voice becomes an economic factor
The Regional Court of Berlin II (case no. 2 O 202/24) recently made a nationally recognised decision on the question of whether and to what extent the right of personality can be violated by the use of artificially generated voices in digital media. In particular, the focus was on how the general right of personality under Article 2 (1) and Article 1 (1) of the German Basic Law should be interpreted in times of artificial intelligence and how far its protection extends when an AI system imitates the characteristic voice of a prominent speaker and is thus used commercially in social networks.
In the underlying case, a prominent German voice actor brought an action against the operator of a YouTube channel who had used an AI-generated imitation of his well-known dubbing voice in two videos. The contributions were satirical in nature and each referred to an affiliated online shop. The plaintiff objected to the unauthorised use of his “voice profile” and demanded compensation and reimbursement of his legal fees. His voice was indeed recognised in the community, which was also reflected in individual comments.
Judicial assessment: The right to one’s own voice
In contrast to many current AI cases, in which copyrights take centre stage, this case is about personal rights – specifically the property right to one’s own voice. The Regional Court of Berlin correctly took the view that the general right of personality also includes the right to one’s own voice, even if the imitation is technical and not human. The decisive factor is whether the impression is created for a significant part of the audience that the person concerned has participated or given their consent. The judges consider this to be an infringement if – as in this case – the voice is specifically associated with a well-known personality and the user has an economic interest.
Consequences: Licence fee and damages for AI voices
The court based the claims on enrichment law claims pursuant to Sections 812 para. 1 sentence 1 case 2, 818 para. 2 BGB, as the AI voice was used for commercial purposes and without consent and the defendant had saved itself a licence fee that it would have had to pay the plaintiff for the use of the voice. Neither freedom of satire nor data protection arguments could justify the interference, as the plaintiff’s economic interest prevailed. Remarkably, the licence fee of €2,000 per video was based on the high recognition factor, the reach of the channel and standard market fees. The plaintiff was reimbursed a total of € 4,000 as well as his legal fees.
Significance for future AI cases
The following also applies to the use of AI-generated voices of real people: The right of personality requires the consent or licence of the person concerned, otherwise there is a risk of claims for damages, among other things.
The judgement makes it clear that synthetic voices are legally protected as strictly as natural voices and that there are clear rules for commercial use. However, future cases may have to be judged differently depending on whether they involve the voices of known persons that are generated and used by AI or voices that – by chance or not – cannot be attributed to known persons.