Denmark’s groundbreaking law on persona rights to combat AI deepfakes

Demark is leading the world against misleading deepfakes. The profusion of AI driven hyper-realistic digital content and  AI slop, now conflicts with prioritizing an individual’s fundamental right to control their own persona. At the forefront of this movement is Denmark, which has recently introduced a sweeping legislative framework to reclassify a person's biometric identity—their face, voice, and physical characteristics—as a form of intellectual property (IP).

Spearheaded by Culture Minister Jakob Engel-Schmidt, the proposed amendments to the national Copyright Act aim to establish the right to one’s own likeness as a core, enforceable asset. This innovative approach seeks to make the unauthorized creation or sharing of AI-generated digital deepfakes illegal, granting every citizen, not just public figures, an unprecedented 50-year protection for their digital identity. In some countries there are personality rights laws, but these usually only apply to celebrities whose likeness has commercial value.

The Danish initiative, expected to be enacted by late 2025 following cross-party support, is a direct response to the accelerating crisis of trust fuelled by synthetic content. Lawmakers recognized that the blurring lines between authentic and fabricated media pose an existential threat to democratic processes and public safety.

The Danish framework is unique in its use of copyright law to combat deepfakes. This strategy is seen by some experts as a more effective tool for enforcing takedown notices on platforms compared to existing, more ambiguous personality or defamation laws.

However, Denmark is not acting in isolation. The global regulatory landscape is rapidly shifting:

• The EU AI Act: As a foundational piece of bloc-wide legislation, the AI Act (in full effect in mid-2025) mandates strict transparency requirements for deepfakes. Providers must clearly and distinguishably label content as artificially generated or manipulated, embedding machine-readable markers (like metadata or watermarks) to ensure traceability.

• The United States: Washington has pursued a more targeted, harm-based approach with the recent passage of the TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act). This federal statute specifically criminalizes the distribution of non-consensual intimate imagery, including AI-generated fakes. Multiple parallel bills, such as the NO FAKES Act (focused on likeness and voice replication) and the Protect Elections from Deceptive AI Act (targeting political misinformation), are pending to broaden federal protections.

• China: By September 2025, new measures will require mandatory technical labeling for all AI-generated or altered content—image, video, audio, and text—ensuring it is visibly marked and traceable within a defined system.

Denmark asserts that it is responding to real-world harms from electoral interference (in Indonesia videos of President Prabowo Subianto speaks in an Instagram video turned out to be deepfakes. Frauds are increasingly now AI-generated likenesses used for sophisticated scams. Brazilian authorities recently dismantled a major fraud ring that used deepfakes of supermodel Gisele Bündchen and other celebrities to promote fake online advertisements, tricking thousands of users out of large sums of money.

A critical and ongoing challenge for regulators is balancing protection of personal rights against freedom of expression. The Danish proposal addresses this by including an explicit carve-out for content that constitutes parody or satire.

While the Danish law leans on the civil right of an individual to control their likeness, it maintains the European Union's broader Digital Services Act (DSA) framework for platform accountability. Under this regime, tech giants face financial penalties if they fail to perform timely removals of flagged content, making them the primary enforcement mechanism against the global distribution of synthetic harms emanating from Denmark.

As Denmark prepares to assume the EU Council presidency, expect its innovative, IP-centric model for deepfake control to be a talking point for Europe.