Knowledge

Speaking to TSF Radio, Dalila Simões, associate in the Digital, Privacy and Cybersecurity practice at TELLES, clarifies some aspects of the recent proposal to amend Danish legislation. According to media reports, the proposal aims to grant citizens copyright over their own image and voice. This would enable them to request the removal of digital content that imitates a real person's voice or image, produced by artificial intelligence, as a way of combating deepfakes.

The measure under discussion stands out for seeking to legally recognise that the unauthorised use of a person's image or voice — even when manipulated by advanced AI technologies — constitutes a violation of their copyright.

In her statements, Dalila Simões clarifies that, contrary to the apparent intention of the Danish proposal – to establish a legal framework for the issue from a copyright perspective – this issue is already regulated in Portugal by the legal regime of personal rights set out in the Constitution of the Portuguese Republic (Article 26), the personality rights set out in the Civil Code (Article 79, right to image), and other legislation. Copyright, on the other hand, refers to rules that aim to protect intellectual creations (e.g. literary and musical works).

TELLES' lawyer also highlights the practical challenges of applying a similar regime in Portugal due to the different nature of personal rights and copyright, adding that "currently, in Portugal, we cannot simply demand that a platform such as Instagram remove an AI-generated video on the basis of copyright, since the image and voice of the person concerned constitute a personality right and not a copyright".

Naturally, the Nordic countries, known for their pragmatism, are likely to use this legal framework of copyright to achieve a practical and quickly applicable solution (namely by notifying platforms to remove such content because it constitutes copyright infringement), a speed that is often incompatible with the delay involved in resorting to legal proceedings or other legal grounds.

This is a high-priority issue, not only because of the potential damage to the target's (the person whose image and voice are being used without consent) good name and reputation, but also because of the potential damage to the recipient of these deepfakes. This content is often used to convey false, fraudulent and criminal messages.

TELLES continues to monitor international legislative developments in this area closely, encouraging critical analysis of the impact of artificial intelligence on fundamental rights.

You can listen to the conversation here [News broadcast on 12 August at 6 p.m., starting at minute 06:10 – Only avalilabe in portuguese].