Just a few seconds of audio are enough to create a deceptively realistic replica of a voice. While this is technically impressive, it raises some legally sensitive questions: Who owns their own voice, and when is its AI-assisted use prohibited?

What is voice cloning?

In a process known as “voice cloning,” artificial intelligence is used to analyze a human voice and then digitally replicate it. Modern AI systems can not only imitate timbre and pitch, but often also speaking rate, intonation, and individual speech patterns. The result is deceptively realistic voice recordings that are virtually indistinguishable from a genuine recording.

This technology is increasingly being used in the media and entertainment industries, for audiobooks, translations, and digital voice assistants. At the same time, however, it can be misused—for example, to create fake voice messages, carry out fraud attempts, or produce deceptively realistic audio deepfakes.

Does a person’s own voice belong to them?

There is currently no independent “right to one’s own voice” under German law. However, this does not mean that voices are unprotected.

The voice is an essential characteristic of one’s personality and may therefore be protected by the general right of personality. Among other things, this right protects a person’s individual identity and is intended to prevent personal characteristics from being used or distorted without consent.

In addition, depending on the specific case, other areas of law may also be relevant. When audio recordings are processed, the data protection requirements of the General Data Protection Regulation (GDPR) generally apply as well. If the recordings involve well-known public figures or professional voice actors, issues related to copyright, competition law, or contract law may also come into play.

What are the risks associated with voice cloning?

The potential applications of voice cloning go far beyond harmless uses. With the help of artificial intelligence, voices can be replicated so realistically that they can be specifically used for fraud or identity theft.

Criminals use AI-generated voices, for example, to deceive family members or company employees. If a person receives a seemingly authentic voice message from a loved one or a company executive, it is often difficult to recognize that it is actually an artificially generated recording. This can result in significant financial losses.

Even in private settings, manipulated audio recordings can be used to disparage people or damage their reputation. Precisely because the recordings often seem authentic, they spread particularly quickly on social media.

When is the use of an AI voice legally problematic?

Whether the use of a cloned voice is permissible depends largely on the circumstances of the individual case.

If a voice is cloned and used without the consent of the person concerned, this may constitute a violation of personal rights. This is particularly true if it gives the impression that the person concerned actually made the statements themselves or endorses certain products, services, or political statements.

If voice recordings are processed for voice cloning, data protection requirements must also be complied with. Companies should therefore carefully assess whether they have an appropriate legal basis or whether the consent of the data subject is required.

Your own voice can have economic value

For many people, the voice is much more than just a personal trait—it’s an integral part of their professional lives. This is especially true for hosts, voice actors, podcasters, musicians, and content creators.

If such a voice is used without consent for advertising, audiobooks, or other commercial purposes, this can have significant economic consequences. Companies should therefore clearly stipulate in a contract, before using AI-generated voices, the extent to which voice recordings may be used and whether they may also be used to train AI systems.

How can those affected protect themselves?

There is currently no way to completely protect against voice cloning. However, those affected can reduce the risk.

Voice recordings should only be published intentionally. Longer, high-quality audio recordings, in particular, make it easier for AI systems to realistically replicate voices. Companies should also establish internal security measures to ensure that sensitive instructions—such as payment authorizations—are not based solely on telephone or voice-based instructions.

If someone’s voice is misused, those affected should act quickly. Depending on the situation, they may be entitled to injunctive relief, rectification, or damages. In some cases, data protection complaints or criminal proceedings may also be considered.

Legal challenges will continue to increase

As artificial intelligence continues to advance at a rapid pace, voice cloning is also becoming increasingly sophisticated. While the technology offers useful applications in many areas, the risk of misuse is growing at the same time.

Companies and individuals should therefore familiarize themselves with the legal framework at an early stage and handle voice recordings responsibly. At the same time, case law is expected to further clarify in the coming years what limits apply to the cloning of human voices and how the general right of personality applies to these technological developments.