The Individual Voice in the Crosshairs of Artificial Intelligence

Whether public figure or private individual – unauthorized access to one’s own voice, its misuse, and the resulting consequences can affect anyone. Such misuse also leads to a general loss of trust and a decrease in the value of audio recordings of all kinds. It is therefore more than welcome that the Berlin Regional Court II (judgment of 20 August 2025, Case No. 2 O 202/24) recently had to address the highly relevant and fascinating question of the right to one’s own voice in the context of AI-generated imitations.

Facts of the Case

In the decided case the parties disputed claims arising from the defendant’s use of an AI-generated voice.

The plaintiff was one of Germany’s best-known voice actors, who had, among other roles, dubbed the popular Hollywood actor Bruce Willis. As a result, the plaintiff’s voice itself enjoys a certain degree of prominence and recognizability, at least in German-speaking countries.

The defendant operated a YouTube channel with almost 200,000 subscribers. On this channel, he published two videos addressing the then-incumbent government. What made the videos special was that they were narrated using a voice generated by artificial intelligence. Since the defendant also ran an online shop, the videos served at least indirectly for advertising purposes.

The plaintiff recognized his own voice in the videos and, through legal counsel, demanded that the defendant ceases using “his” voice and pay damages as well as reimburse the attorney’s fees for the warning letter. Although the defendant did sign the requested cease-and-desist declaration, he refused to pay the requested licensing damages of € 2,000 per video – reflecting the plaintiff’s usual fee practice – or to reimburse the out-of-court legal fees. He argued that an AI-generated voice, despite its similarity, was merely an imitation, not the plaintiff’s real voice.

Reasons for the Decision

The Berlin Regional Court II ordered the defendant to pay damages and reimburse the legal fees, because he had infringed the plaintiff’s right to his own voice, which is assigned to the plaintiff alone, including its economic components:

• As recognized in case law and legal literature, the general right of personality (Art. 2(1) in conjunction with Art. 1(1) of the German Constitution) includes the right to one’s own voice. The extent to which the use of a well-known voice for advertising purposes affects the right of personality is comparable to the use of a person’s image or name. Referring to the Federal Court of Justice (BGH), the Regional Court further explained that a person’s picture, name, and other personality traits can have substantial economic value in individual cases. Therefore, only the rights holder may decide whether and under what conditions any of these personality attributes – including the voice – may be used for the commercial interests of third parties.

• The use of an AI-generated imitation of a voice in a video also interferes with the affected person’s right of personality. The artificially generated imitation of the plaintiff’s real voice is legally equivalent to an imitation created by a human voice actor. What matters is the confusion created by the deliberate similarity, leading viewers to believe that the plaintiff, as the German voice of Bruce Willis, had allowed his voice to be used for the defendant’s videos.

• In principle, the use of a voice may be justified by analogy to the provisions on the use of images (§§ 22, 23 German Art Copyright Act (KUG)), but the necessary requirements (such as predominantly satirical use) were not met. The court therefore ordered the defendant to pay a hypothetical license fee of € 2,000 per video and to reimburse the out-of-court attorney’s fees. The court held that case law concerning the economic value of unauthorized use of images for commercial purposes applies equally to the commercial use of someone’s voice – including AI-generated imitations. Anyone who gains a commercial advantage by using a well-known voice, whether that of a celebrity or of their dubbing actor, must bear the resulting economic consequences.

Practical Note

In this groundbreaking decision the Regional Court addressed the key legal issues surrounding the use of AI-generated voices with compelling reasoning.

First, absolutely agreeable is the court’s classification of the right to one’s own voice as part of the constitutionally protected general right of personality, including its economic implications.

However, the real core and value of the decision lies in the judicial finding that AI-generated voice imitations also interfere with the affected person’s right of personality and are therefore generally prohibited. This is a crucial step toward effectively combating unauthorized AI access to a person’s voice. This applies not only to cases like the present one, where prominent voices are used for advertising or attention-grabbing purposes.

Nevertheless, the Berlin Regional Court II only briefly touched on a non-trivial issue – also relevant in this case: To whom is the voice of a dubbing actor attributed – the voice actor, the dubbed actor, or the fictional character (here, e.g., John McClane from Die Hard)? The court merely noted that the public knows the actor is not German and is dubbed and therefore attributes the voice to the plaintiff. It remains to be seen whether cases arise in which viewers believe the actor is speaking or the commercial value of the fictional character is being exploited. This, however, does not change the fact that using the voice of the dubbing actor infringes the actor’s rights.

Also relevant for practice is the question of the extent to which the justification grounds under § 23(1) KUG – particularly uses for contemporary events (No. 1) and artistic purposes (No. 4) – allow such use. Case law on the artistic use of images of well-known personalities in advertising is quite generous, as illustrated by the BGH ruling of 26 October 2006 (Case No. I ZR 182/04) in the Sixt advertising case. Applying this standard to voice use, it is conceivable that the commercial opportunities for actors and voice actors could be impaired.

Beyond that, the use of someone’s voice through AI can potentially affect anyone. This relatively new phenomenon can also play a role in very private contexts, with significant consequences. Because AI can easily generate voice messages or audio files in virtually any voice, AI voice imitation can cause considerable harm depending on the circumstances – not only economically, but also in a person’s private life.

The problem affects not only individuals but also society as a whole. The danger posed by “AI voices” introduces substantial uncertainty into business and legal transactions. Also the administration of justice is not immune: The evidentiary value of audio recordings in both civil and criminal proceedings will inevitably be weakened. What was once a fairly reliable form of evidence must now be scrutinized much more critically.

Given these far-reaching and serious consequences associated with AI-generated voices, a clear commitment to the (civil-law) protection of one’s own voice as an expression of the general right of personality – as the Berlin Regional Court II has now articulated – is particularly valuable.

A key future question will be whether, in addition to civil law, criminal law should also address the unauthorized use of AI-generated voices. There is currently no such criminal provision, as Section 201a of the German Criminal Code (StGB) only concerns unauthorized image recordings. Considering the importance of audio recordings for the public and especially for the administration of justice, such a provision seems to be at least worth discussing.