Europe draws a red line on AI nudifier apps

A hand holding a phone with AI apps on its screen, beside a photograph of Elon Musk

In a landmark vote, European Parliament committees moved to ban AI tools that generate non-consensual sexual images of real people. It is not yet law, but the political signal is unmistakable.

Video: https://www.facebook.com/reel/1330164925589847

What was actually voted on

On 18 March 2026, the Internal Market (IMCO) and Civil Liberties (LIBE) committees of the European Parliament adopted their joint position on a proposal amending the EU’s Artificial Intelligence Act, by 101 votes in favour, 9 against, and 8 abstentions.

The headline measure: a proposed ban on so-called “nudifier” systems, meaning AI tools that create or manipulate images to depict an identifiable real person in a sexually explicit or intimate way without their consent. The ban would not apply to systems that have effective safety measures in place to prevent such content from being generated.

The context for the vote was stark. According to data from the NGO Centre for Countering Digital Hate, Grok, the AI tool integrated into Elon Musk’s platform X, generated 3 million sexually explicit images and 20,000 artificial reproductions of child sexual abuse material over just eleven days between late 2025 and early 2026.

Co-rapporteur Michael McNamara (Renew, Ireland) put it plainly: “AI must never be used to humiliate, exploit or endanger people. This is something our citizens expect of us.”

What else was in the vote

The nudifier ban was part of a broader “Digital Omnibus” simplification package. Other key elements include:

  • Deadlines for high-risk AI systems (covering biometrics, critical infrastructure, education, employment, law enforcement, justice, and border management) pushed back to 2 December 2027, because key technical standards will not be finalised in time.

  • An extension of the deadline for watermarking rules on AI-generated audio, image, video, and text content, to 2 November 2026, shorter than the European Commission’s proposed February 2027.

  • Extension of SME-style regulatory reliefs to small mid-cap companies with up to 750 employees and €150 million turnover.

How does it become law?

This is the critical point: it is not law yet. The legislative path ahead has several steps.

  1. Committee vote (done). IMCO and LIBE adopted their joint position on 18 March 2026.

  2. Full Parliament plenary vote. Expected on 26 March 2026. This gives Parliament its formal negotiating mandate.

  3. Trilogue negotiations. Parliament and the Council of the EU (representing member state governments) negotiate the final text. This process typically takes several months.

  4. Formal adoption. Both Parliament and Council formally adopt the agreed text.

  5. Entry into force. As an EU Regulation (not a Directive), it will apply directly across all 27 member states with no need for national transposition legislation.

Definitive adoption is not expected before mid-2026 at the earliest.

How will it be enforced?

The AI Act designates the EU AI Office, based within the European Commission, as the central supervisory authority for general-purpose AI models. Alongside this, each member state must designate its own national competent authority.

For banned practices such as nudifier apps, enforcement would likely target both the developers and providers of the tools and the platforms that host or distribute them.

The penalties are significant. Violations of prohibited practices provisions, the most serious category, carry fines of up to €35 million or 7% of global annual turnover, whichever is higher.

The bottom line

The committee vote is a meaningful milestone and a clear political signal. But the ban is not yet law. The plenary vote on 26 March will determine whether Parliament has a formal mandate to negotiate, after which the real work of shaping the final text begins.

For women and girls across Europe, this vote matters. Non-consensual intimate imagery causes profound harm, and the scale revealed by the Grok data shows just how rapidly AI is amplifying that threat. The EU is moving to address it. Whether the final law will have the teeth to do so effectively depends on what survives the trilogue.

Sources: European Parliament press release · RTE News · Trending Topics · EUNews · AiActo

This post was prepared for the EUWMA blog. For personal or community sharing only.

Would you like to reply? You can do so in our Facebook group, where Facebook will ask you to join the group first, or write to us at [email protected].

Back to Blog/News