The rise of AI and its legal implications
In recent years, technological advancements have brought about impressive yet controversial achievements. For instance, an AI-generated cover of the song "Saiyan" was released without Angèle's consent, sparking a debate on the ethics of such technology. Similarly, Scarlett Johansson discovered that the voice of a virtual assistant bore an uncanny resemblance to hers, despite her refusal to license her voice for this purpose. On the other hand, AI has been used positively, such as when David Beckham conveyed his message in nine different languages as part of the Malaria No More campaign. While AI offers unprecedented possibilities, it also raises significant legal questions.
The use of AI to replicate someone's voice or image for commercial, artistic or informational purposes demands careful consideration during contractual negotiations. While such use might sometimes fall under the umbrella of free expression, it can also cause significant harm and become a tool for misinformation. In Belgium, although AI-specific legislation is still in development, existing legal frameworks – such as copyright law, image rights and privacy law – regulate these practices. This article focuses on AI in the context of personality rights.
Understanding personality rights
Personality rights are subjective rights granted to every individual simply by virtue of being a person, aimed at safeguarding the physical, psychological and moral integrity of their personality. These rights include image rights, voice rights, the right to privacy and the right to protect one’s reputation. They serve to protect individual attributes from encroachments by others or the authorities.
Image rights
Image rights allow individuals to prevent the publication of their likeness without their consent, provided they are recognizable, regardless of the medium used (photo, video, sculpture, etc.). This right is violated whenever someone’s image is used without authorization or when an authorized image is repurposed for unauthorized uses.
However, due to their status, public figures – such as movie stars, top athletes and influencers – may have to tolerate the dissemination of their image without explicit consent. This exception is grounded in the broad interpretation of the right to information. Nevertheless, this right to information must be balanced with respect for privacy, honour and reputation, and should not involve unauthorized commercial use. The use of their image must be related to their role or the reason they are of public interest.
Voice rights
Similar to image rights, voice rights allow an individual to oppose the unauthorized use of their voice.
Right to privacy
The right to privacy ensures that individuals can live free from undue intrusion and can control what aspects of their private lives they choose to disclose. This includes the protection of personal data and image rights. The GDPR is the main legal instrument regulating and restricting the processing of personal data.
Right to reputation and honour
A person's dignity is also protected under personality rights. Deepfakes, for example, can be used to create misleading videos or audio recordings that harm someone’s reputation by spreading false information. Thus, generating a video with synthesized images of a person in a degrading situation could constitute an infringement of their reputation and honour.
Conclusion: ensuring compliance when using AI
When creating content using AI, it is crucial to respect personality rights. This requires obtaining the consent of individuals whose voice or image is being used. Such consent should cover several aspects: the context in which the generated image or voice will be used, the purpose (commercial, artistic, informational, etc.), distribution channels, duration of use, geographical scope, remuneration and whether the individual can monetize their image or voice with competitors. If you outsource the generation of this content, it is essential to communicate these limitations clearly to avoid any violation of the concerned person’s rights.
Moreover, Regulation 2024/1689 of 13 June 2024, establishing harmonized rules regarding AI ("AI Act"), introduces specific provisions concerning deepfakes. Starting 2 August 2026, the AI Act mandates transparency for AI systems generating deepfakes (Article 50). This means that any AI-generated creation, such as a deepfake, must clearly state that it was generated or manipulated artificially.
This evolving legal landscape highlights the importance of understanding and respecting personality rights in the age of AI, ensuring that technological advancements do not come at the expense of individuals’ rights and dignity.
This article was written by Elisa Le Grelle, with the valued assistance of ChatGPT. It is part of the series “AI and intellectual property rights”, written by the IP lawyers at CMS in Belgium.
