Denmark’s culture minister, Zenia Stampe, put a bill before the Folketing on 8 October banning publication of realistic digital imitations of a person’s appearance, voice and movements without consent. She proposes effect from 1 January 2027. It is the second attempt: the prior version lapsed in Danish elections, after the 31 March 2026 and 1 July deadlines were missed. Stampe says she “espera” it can start “ved årsskiftet,” at the turn of the year.
- Key Point: No separate deepfake crime is created. Denmark amends its copyright act, the ophavsretsloven, so identity enters via authors’ rights.
- Key Point: The general rule covers all natural persons, foreigners included, for 50 years from the end of the death year. EEA performers get a separate, collectively managed right against imitations of performances, improvisations included.
- Key Point: Dissemination is regulated, not creation; private use generally stays out. Intent is unnecessary, so negligent publication counts, and AI plus ordinary digital editing are covered. Publishers must prove consent, revocable anytime without formalities for future releases; contracts before 1 January 2027 are untouched.
- Key Point: Remedies are civil: damages and non-pecuniary compensation, plus amended article 84 for court-ordered removal. There is no criminal offence or penal sanction.
Identity inside copyright
The copyright route is deliberate. The broad provision reaches every natural person, including non-Danes, for half a century after the year of death—the term used for a work. EEA performers gain a separate collectively manageable right covering imitations of their performances, improvisations included.
The bill governs dissemination, not making; private use normally falls outside. No intent is needed, so careless publication is caught, and both AI and conventional editing apply. The publisher must show consent; consent can be withdrawn at any time without formality, though only for future publications. Contracts signed before 1 January 2027 survive.
Platform promises and objections
Enforcement is civil: damages and non-pecuniary compensation, and amended article 84 lets a court order removal from the medium. The sharpest contradiction is platforms. Stampe calls fast removal by big platforms a central aim:
“techgiganterne til at pille det manipulerede indhold ned hurtigt”
But the text sets no deadline, defines no request, and requires no prior filtering. It gives people a legal basis to demand removal, not a monitoring duty. Fast removal is political, not binding. A technical issue remains: Digital Services Act hosting services are excluded from “medium,” yet social networks are such services; clarification is expected in committee.
Exceptions include caricature, parody, pastiche, satire and criticism of power and society, with an open “and similar” list. They fail where imitation is disinformation capable of grave danger to essential rights or interests, or concrete danger to life or property. Fantastical creatures and clearly stylised portraits are outside scope.
Criticism comes from both sides, but dates from February 2026 and targeted the earlier TRIS notification to the European Commission; no post-8 October organisational reactions were found. Justitia says copyright is inadequate versus criminal law and satire may be too narrow:
“selv tydelige karikaturer principielt vil kunne censureres”
It warns powerful politicians could remove satire; this was its first TRIS response. VideoGamesEurope—the European video-game trade body whose members include Microsoft, Epic Games, Ubisoft, Netflix and Sony—claims internal-market breaches and parallel safeguards to GDPR, the Digital Services Act and AI Act guidelines:
“Dette tvinger udviklerne til en evig risikovurdering af hver karaktermodel, animation og stemmereplik”
Rettighedsalliancen seeks clearer definitions of realistic imitation and confusion risk, warning an AI-generated label could remove protection; Danske Advokater supports.
Spain takes another route
For Spain, less changes than it seems: Madrid is moving another way. On 7 July 2026 the Council of Ministers approved an Organic Law bill on civil protection of honour, personal and family privacy and one’s own image. It replaces the 1982 law and, for the first time, declares illegitimate unauthorised use of a person’s voice or image for advertising or commercial purposes through AI. Félix Bolaños, minister of the presidency, justice and relations with the Cortes, said:
“Devolvemos a las personas el control de sus propias fotos, vídeos y contenidos”
The contrast is technical: Spain protects image through honour and privacy; Denmark through copyright.
Denmark’s design is clever but brittle. Copyright supplies ownership, duration and courts, so grafting likeness onto it avoids a new regime. Yet identity becomes a 50-year post-mortem asset, and civil remedies burden those able to sue. A victim needs a court win, not an authority acting alone—a poor fit when fakes spread fast.