AI nude images run into slow enforcement
The Netherlands and the European Union already ban many forms of abuse, but victims face problems with distribution and oversight.
The rapid spread of AI-generated nude images has exposed a legal problem: more and more rules are being introduced, but removal and enforcement remain difficult. Dutch ministers say unsolicited deepfakes are largely already criminal or unlawful.
The police warn that synthetic media are becoming cheaper and easier to create. So-called deepnudes can place someone's face, without their consent, on sexual material. The police treat this as a form of abuse and refer victims to help with removal and reporting the offence.
In answers to parliamentary questions in the Lower House (Tweede Kamer), the government wrote that existing Dutch criminal and civil law can already address many of these situations. The GDPR and portrait rights may also play a role when identifiable personal data or an image of someone are processed and distributed without a valid legal basis.
That does not mean that every image is removed immediately or that every creator is tracked down. Distribution may run through multiple platforms, private groups and foreign providers. In July, the police reported that servers in the Netherlands had been seized in an international investigation into the Motherless website; there had been no arrests at that point.
Online platforms also have obligations. According to the Dutch government's responses, hosting services must be able to receive reports of illegal content and assess them promptly. Large platforms must also assess and mitigate systemic risks, including the spread of manipulated sexually explicit images.
In its evaluation of the AI Act, the European Commission previously identified a possible gap concerning systems that generate non-consensual intimate content or child sexual abuse material. At the same time, the Commission refers to a political agreement from May 2026 that includes a ban on such systems. The precise effect depends on further legislation and enforcement.
The claim that AI pornography has ‘no limits’ is therefore too absolute. There are legal limits, but victims do not always experience them as protection when images quickly reappear, providers are beyond reach or evidence disappears. The debate is therefore not only about new bans, but also about reporting channels, platform design, international cooperation and the capacity of regulators and the police.
One story, several perspectives
What is established
- Dutch authorities say that many non-consensual deepnudes are already criminal or unlawful.
- Platforms must be able to receive and handle reports of illegal content.
- The European Commission described a possible regulatory gap and referred to a political agreement on expanding the bans.
Left
Arguments Victims are entitled to rapid removal, active investigations and strong protection from platform algorithms that further spread harmful images. Self-regulation alone is insufficient.
Values Bodily autonomy, equality and protection from digital abuse of power.
Consequences Stronger platform obligations and better assistance could limit harm, but require international cooperation and public enforcement capacity.
Centre
Arguments Existing criminal and privacy rules should first be enforced consistently. New bans may help, but must be precise so that platforms can tackle illegal content without blocking legitimate imagery or satire.
Values Legal certainty, proportionality and effective implementation.
Consequences A combination of clear reporting channels, preservation of evidence and oversight could improve protection without creating an unworkable blanket internet ban.
Right
Arguments The creator and distributor of the abuse should be the focus; platforms should respond when they know about illegal content, but should not be held responsible for every image users upload. Overly broad rules could harm innovation and freedom of expression.
Values Personal responsibility, property rights and restraint in government intervention.
Consequences Targeted enforcement may work more quickly, but victims could continue to fall between providers and jurisdictions if platform responsibility is interpreted too narrowly.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The text distinguishes between existing legal options, announced European policy changes and practical enforcement. It contains no explicit sexual details or unsubstantiated figures on victims.
- confirmed The police say that deepfakes and deepnudes are becoming easier and cheaper to create. — Stated on the police information page. source
- confirmed The government says that unsolicited deepnudes are largely criminal or unlawful under existing Dutch law. — Mentioned in the official answers to parliamentary questions. source
- confirmed In July, the police seized Motherless servers and reported that there had been no arrests at the time. — Mentioned in the police statement. source
- confirmed The European Commission described a possible gap concerning AI systems for non-consensual intimate content. — Stated in the report on prohibited AI practices. source
Editor's note
The legal frameworks and police actions can be publicly verified. The scale of the problem and the effectiveness of future EU rules cannot be fully established from the available sources.Sources
- Wat is deepfake of deepnude? — Politie
- Kamervragen over AI-uitkleedsoftware — Overheid.nl
- Report on the review of prohibitions and high-risk AI — Europese Commissie
- Doorzoekingen in onderzoek naar website Motherless — Politie
More on this in Dutch media
- RTL Nieuws — „deepfake deepnudes”
- NOS — „deepfake deepnudes”
- Het Parool — „deepfake deepnudes”