By registering a sound trademark to protect his iconic phrase “Alright, alright, alright”, the actor Matthew McConaughey has chosen a striking legal route to defend himself against unauthorised use of his voice by AI tools, says Rosalie Huijgen.
Over the past few months, actor Matthew McConaughey has obtained a series of trademark registrations from the United States Patent and Trademark Office (USPTO), including a sound trademark for his iconic phrase: “Alright, alright, alright”. The phrase, first uttered in the 1993 film Dazed and Confused, has since become inextricably linked to his public persona.
What's notable about this registration is that it applies not only to the words themselves, but to the specific way McConaughey pronounces them: his intonation, rhythm and tone of voice. Equally remarkably, the filings provide a striking legal route to defend himself against unauthorised use of his voice by artificial intelligence (AI). His decision to seek trademark protection raises a fundamental question: can trademark law be an effective weapon against AI deepfakes and voice impersonations?
AI deepfakes: A growing problem
McConaughey's trademark filings are no coincidence. Today’s AI technology can realistically simulate voices. Even with limited training data, a model can generate a voice that is virtually indistinguishable from the real thing. This poses significant risks for celebrities, as well as for those not publicly known.
For example, voices could be used without permission in advertisements, other commercial expressions or political messages, leading to reputational damage, public deception and loss of commercial value for the person in question. The problem is that the existing legal framework currently offers only limited protection against such practices.
Why classic grounds for protection fall short
Taking EU legislation as an example, the following provisions do not provide a viable means of protection:
Although you can recognise someone by their voice, portrait rights only protect the outward appearance of a person. Legally speaking, a voice is not a portrait and therefore does not fall under this type of legal protection.
Copyright also offers little security. Protection is reserved for creative works with original character and the personal stamp of the creator. Generally, a voice or short utterance does not meet this requirement. Moreover, copyright enforcement usually requires the reproduction of an existing work. AI deepfakes are often new creations that merely resemble an existing voice, without copying a specific recording.
Several countries are exploring solutions to offer copyright protection for voices. In Denmark, a bill is being proposed to grant people copyright over their own body, facial features and voice. Similar proposals are also being developed in the Netherlands, but these require a basis within the Neighbouring Rights Act.
- European Fundamental Rights, GDPR and AI Act
The shortcomings of classical IP rights do not mean that voices are completely unprotected. Other protection mechanisms can be invoked within European law.
For example, the unauthorised use of a voice can violate the right to respect for private life (Article 8 of the European Convention on Human Rights, ECHR). However, this is offset by fundamental rights of the person using the voice, such as freedom of expression (Article 10, ECHR) and artistic freedom (Article 13, EU Charter). This requires a balancing of interests: only a sufficiently compelling interest can justify restricting the latter freedoms.
In addition, the EU General Data Protection Regulation (GDPR) may be relevant, as a voice can be considered biometric personal data. Processing such data is, in principle, prohibited without a valid legal basis. However, the GDPR includes broad exceptions for journalistic, artistic and literary purposes, meaning that this protection is often limited in practice.
The European AI Act also offers insufficient protection. The regulation contains rules for transparency and labelling of AI-generated content, but does not explicitly regulate consent or control over the use of voices.
Trademark law as an alternative route
Against this backdrop, McConaughey's choice to use trademark law is easy to understand. By registering his distinctive utterance as a sound trademark, he places his voice within a relatively strict and technical legal framework. The question thus shifts from "Is someone allowed to use my voice?" to "Is my trademark being (unlawfully) used?"
Sound trademark infringement occurs when a third party uses a (nearly) identical sound for similar products or services without permission. The assessment focuses on the overall impression, with aural similarity (how the sound sounds) being central.
This should then lead to a likelihood of confusion among the public. The public doesn't necessarily have to have the impression that McConaughey recorded the message himself; it can be sufficient that the impression is created that the voice originates from him or reminds them of the registered trademark.
How powerful is a sound trademark?
The major advantage of trademark law is its enforceability. Unlike privacy or fundamental rights, no explicit balancing of interests is required. At the same time, trademark law offers broad protection: the specific, registered sound is protected for the goods and services for which the trademark is registered and is used in accordance with the obligation of use and for similar goods and services.
Before a trademark owner can derive rights from a sound trademark, it must, of course, first be registered. The use of a human voice as a sound trademark is, in principle, possible, provided the statutory criteria for a sound trademark are met. First and foremost, the voice must possess sufficient distinctiveness. It must enable consumers to recognise and distinguish the goods and/or services of one company from those of others. A well-known voice, such as that of Matthew McConaughey, can satisfy this requirement provided the public genuinely associates the voice with him or with a specific company.
Furthermore, the voice must be capable of being reproduced clearly and accurately as a sound trademark, for example, in the form of an audio file. In practice, however, registering a voice as a sound trademark can be difficult. A voice is generally considered a natural and general means of communication. If the voice doesn't deviate sufficiently from common voice sounds, the required distinctiveness is lacking. Even with well-known voices, it's not self-evident that the public will recognise them as a trademark. This lack of distinctiveness can hinder registration.
If a sound trademark – such as "Alright, alright, alright" – does meet all the requirements and is successfully registered, the ability to enforce it depends largely on how judges assess the key questions: is there trademark use, is there a likelihood of confusion, and does the unauthorised use of the voice (e.g. in an AI deepfake) fall within the trademark description and relevant goods and services?
In addition, AI raises new questions about liability and ultimate enforcement: who and/or which entity is the infringer, and can you be held liable for the unlawful use of the trademark? The developer of the design, the user, or the platform that facilitates the voice? Trademark law doesn't provide a clear answer to this question.
Turning a trademark voice into a sound trademark
Registering a voice as a sound trademark is a smart and innovative move in the fight against AI deepfakes. It offers an additional legal weapon for when traditional grounds for protection fall short. At the same time, it's not a watertight shield. Its effectiveness depends largely on how judges assess the key questions: is there trademark use, is there a likelihood of confusion and does the AI deepfake fall within the trademark description?
In conclusion, McConaughey has broadened the legal playing field, but only time will tell the extent to which trademark law is truly capable of effectively protecting a sound trademark against unauthorised use by AI.
To find out more about sound trademarks, including the key steps to take to trademark a sound, please speak to your Novagraaf attorney or contact our team.
Close
