Denmark Moves Toward AI Copyright Rules for Voice and Appearance

Denmark is taking a ground-breaking step in the fight against deepfake technology, proposing an amendment to its copyright laws that would allow individuals to own their own likeness. This innovative approach, backed by a broad consensus in the Danish Parliament, aims to empower citizens to demand the removal of unauthorised digital reproductions of their appearance and voice from online platforms.

Deepfakes are incredibly realistic digital fabrications of a person’s image, voice, or actions and are becoming more sophisticated and harder to distinguish from reality, posing significant risks, including harassment, coercion, and even falsely implicating individuals in crimes.

Current legal frameworks, like the United States’ Digital Millennium Copyright Act (DMCA), typically protect creative works like photos or recordings, but not a person’s inherent identity. This means only individuals with significant resources, like Scarlett Johansson, can effectively fight against deepfakes since ordinary citizens often lack the means to pursue legal action.

For your information, in 2024, Scarlett Johansson and OpenAI faced a controversy when OpenAI released a new voice for its ChatGPT system, “Sky,” resembling her voice. Johansson declined an offer from OpenAI’s CEO, Sam Altman, to license her voice, still the voice was still launched. Johansson was shocked and threatened legal action, leading to OpenAI halting Sky’s voice.

Similarly, viral videos showcasing a DeepTomCruise created by Metaphysic demonstrated just how convincing and unsettling deepfake technology had become.

Adding to these concerns, Hackread recently reported a sophisticated AI-powered phishing operation targeting YouTube content creators. Scammers used a highly realistic deepfake video of YouTube’s CEO, Neal Mohan, to announce fake changes to monetisation policies, aiming to steal login credentials and install malware.

Danish Culture Minister Jakob Engel-Schmidt has strongly voiced his concern, stating that “human beings can be run through the digital copy machine and be misused for all sorts of purposes, and I’m not willing to accept that.” He highlighted that existing laws are insufficient to protect individuals from generative AI.

The proposed Danish law would legally define unauthorised digital representations, specifically targeting “very realistic digital representations of a person, including their appearance and voice,” thereby providing a direct legal basis for removal requests and potential financial compensation for victims. It includes exceptions for parody and satire to protect freedom of speech. The legislative proposal is undergoing public review before formal submission this autumn.

Denmark intends to use its upcoming EU presidency to champion a new approach to deepfake regulation across Europe. Unlike current efforts that largely penalise harmful deepfake uses after they occur, Denmark proposes granting individuals copyright over their digital likeness to prevent misuse from the outset.

Apart from Denmark, companies like Metaphysic are exploring alternative solutions. Metaphysic is working on a system that allows individuals to create an AI-generated avatar of themselves and copyright it, which would fall under existing copyright protections, offering a potential workaround in countries where personal likeness isn’t directly copyrightable. However, Denmark’s proposal is a more straightforward solution to a pressing global challenge, and its success could influence future deepfake legislation worldwide.