Denmark is preparing to become one of the first countries in Europe to give citizens legal control over their own face and voice in the digital world, just as new UNICEF research shows at least 1.2 million children globally had their images manipulated into sexually explicit AI deepfakes in the past year.
The scale of the problem is staggering. According to UNICEF, in some countries roughly one in 25 children has experienced this abuse. That means in a typical classroom, one child has been a victim.
Denmark is not among the 11 countries surveyed, which include Mexico, Pakistan, Serbia, Morocco and Montenegro. But the findings matter here because they show this is no longer limited to celebrity scandals or political misinformation. It has become a global child protection crisis, and Danish lawmakers are trying to get ahead of it.
A Right to Your Own Body and Voice
Denmark’s response is bold and legally unusual. The government struck a broad political agreement in mid 2025 to restrict the sharing of realistic AI-generated imitations of people’s faces and voices. The law is expected to take effect sometime in 2026, though the exact date remains unclear.
What makes the Danish approach distinctive is that it uses copyright law as the main vehicle. The idea is to give every person a legal claim over digital reproductions of their personal characteristics. If someone shares a realistic deepfake of you without consent, you can demand platforms take it down and potentially seek compensation.
Several European legal analysts have questioned whether copyright is the right framework for protecting personality rights. But Danish authorities argue it offers a clear and familiar enforcement route, especially against platforms that are slow to remove abusive content.
Where the Law Draws the Line
The draft framework explicitly protects satire, parody and caricature. That carve-out is politically necessary because any deepfake law risks chilling free expression. But it also creates a gray zone. A realistic fake can be labeled satire while still causing real harm.
I have watched Denmark wrestle with balancing free speech and online safety for years, particularly around social media regulation. This is the same tension, only sharper. The law tries to draw a bright line, but courts will have to figure out where parody ends and abuse begins.
The law is also designed to work with the EU Digital Services Act. Platforms will face economic consequences if they fail to remove unlawful deepfakes after notice. That matters because deepfakes spread fast and get re-uploaded constantly. Speed is everything.
The Child Protection Gap
UNICEF’s framing is unambiguous. As stated by Line Grove Hermansen, director for communication and advocacy at UNICEF Denmark, deepfake abuse is abuse. There is nothing fake about the harm it causes.
UNICEF is urging governments to expand definitions of child sexual abuse material so they explicitly include AI-generated content. It also wants production, possession and distribution criminalized. That goes further than Denmark’s current copyright-based proposal, which focuses on takedown and civil remedies rather than criminal prosecution.
The gap is real. Many legal definitions were written before generative AI became cheap and widely accessible. If AI-generated child sexual abuse material is not clearly covered, prosecutors may struggle to act. Denmark’s new law might help victims demand removal, but it does not necessarily close the criminal enforcement gap.
Search Engines Are Part of the Problem
A report from Danish advocacy group Digitalt Ansvar claims that search engines generate between 43 and 83 percent of traffic to sites hosting manipulated sexual content. The group specifically names Google as still directing users toward this material.
That finding is significant because it suggests platform moderation alone is not enough. If search engines keep surfacing abusive content prominently, removal becomes a game of whack-a-mole. The Danish law gives individuals a tool to fight back, but enforcement will depend on how quickly platforms and search engines respond.
A Model for Europe?
European commentary describes Denmark’s proposal as one of the most far-reaching attempts yet to regulate AI-generated identity abuse. The law will likely apply only in Denmark, but it could influence other EU policymakers.
The practical challenge is that deepfakes are borderless. Danish law can protect users at home, but it has limited reach against creators and platforms operating abroad. That makes this a test case for whether national regulation can tackle a global problem.
UNICEF says AI developers must implement safety by design and build in robust safeguards. Some are doing that. Many are not. The risk grows when generative AI tools are integrated directly into social media, where manipulated images can go viral in minutes.
Denmark is trying to act before the technology runs ahead of the law. Whether this model works will depend on enforcement, platform cooperation, and whether other countries follow.
Sources and References
UNICEF: Deepfake-overgreb er overgreb
The Danish Dream: Danish ex-minister charged with child abuse material
The Danish Dream: Leading
