Core Summary

Reuters reported that Denmark plans to prohibit sharing AI-generated deepfakes without the consent of the person depicted. The proposal would place consent, platform duties and judicial evidence rules in one framework as Europe moves generative-AI governance beyond model safety toward protection of personal image rights.

Event Details

  • Legislative direction: The Danish government reportedly wants consent before distributing clearly misleading synthetic imagery, rather than relying only on a label stating that the content was AI-generated.
  • Platform duties: Disputes will force platforms to answer who must remove content, how long records should be retained and how original material can be supplied to police and courts.
  • Evidence challenge: Deepfakes are becoming shorter and more realistic, so ordinary users cannot reliably judge authenticity by sight alone; evidence rules must balance speed and accuracy.
  • European context: The EU has already adopted a broader AI regulatory framework, but member states still differ in how they apply image rights, privacy, media liability and criminal responsibility.

Panoramic Analysis

At first glance, the proposal limits the spread of an image or a video. In substance, it redistributes rights in the digital environment. Image rights traditionally governed advertising, news coverage and commercial exploitation without permission. Generative AI turns that question into a sequence: can a system copy a person, circulate the copy and explain it afterward? If the system relies only on removal after harm, the victim still bears the burden of proof. Writing prior consent into the rule can establish a clearer starting point for responsibility.

Platforms will face a substantial increase in compliance work. They may need to preserve publication records, model-call logs and authorization credentials while distinguishing satire, public-interest footage and malicious synthetic media. Blanket removal would damage journalism and artistic expression; doing nothing would magnify abuse. Denmark’s proposal will therefore need risk tiers, appeal channels and precise allocation of evidentiary responsibility, not only a prohibition.

For the public, the most visible change may be a new default: do not forward. If a distributor cannot demonstrate consent, the risk should rise. This could change social-media habits and make people identify synthetic content earlier. Over time, European regulatory competition may shift from who owns the strongest models to who can make personal control executable in digital life.

Comparative Perspectives

  • Danish government: Emphasizes personal dignity and reducing the burden of proving harm, rather than leaving victims to investigate alone.
  • Platform companies: Support common rules but warn about compliance costs, accidental removal of legitimate content and inconsistent national definitions.
  • Digital-rights groups: Back an explicit consent principle while warning against letting large platforms monopolize verification and takedown systems.
  • Media and researchers: Seek room for journalism, satire and scientific detection, favoring risk- and context-based treatment.
  • Legal scholars: Note that countries have different traditions in image rights, making the result depend on how Danish law defines consent and misleading content.

Editor: GoodInfo Global News Team