How the EU is tackling pornographic deepfakes: a breakdown

The short version

The European Parliament has voted to ban tools that generate pornographic deepfakes. This measure targets the creation of non-consensual nude images using artificial intelligence. It is not a retroactive erasure of existing content, nor is it a general ban on all AI image generators.

According to a report by 20 Minutes, this vote represents a specific legislative response to a surge in synthetic media. The mechanism relies on a revision of the EU’s AI Act, originally adopted in 2024. The goal is to criminalize the production tools themselves, rather than just punishing individual users after the fact.

This guide decodes the legislative logic, the scale of the reported problem, and the existing tools for victims. It separates the confirmed vote from the ongoing development of criminal penalties.

How it works

The core mechanism is a shift from policing content to policing tools. Under the current framework, authorities often react after images are published. The new approach, as reported by 20 Minutes, aims to stop the creation process.

The European Parliament’s vote targets the software and algorithms capable of generating these specific types of images. By banning the tools, the law seeks to remove the means of production. This is distinct from a content moderation policy, which relies on platforms removing images after they are uploaded.

This legislative move is embedded within a revision of the AI Act. The original law, adopted in 2024, established a baseline for AI regulation. The current update tightens these rules specifically for non-consensual sexual content.

The scope covers a wide range of targets. As noted in the reporting, no public figure is immune. The phenomenon affects everyone from global pop stars like Taylor Swift to political leaders such as Giorgia Meloni. The mechanism does not distinguish between celebrities and private citizens; the illegality lies in the non-consensual nature of the generated image.

What is sourced

The scale of the issue driving this vote is significant. According to 20 Minutes, a specific incident in January 2026 highlighted the urgency.

During that month, the AI model Grok, developed by Elon Musk, reportedly enabled the creation of three million pornographic deepfake contents in just a few days. This volume triggered investigations across the European Union.

The legislative response is a direct reaction to this volume. The European Parliament voted to prohibit these tools. The source material attributes this vote to the need to address the multiplication of nude images generated by AI.

Furthermore, the reporting indicates that the European Union is currently working on a bill to criminalize these deepfakes. While the tool ban is a step, the full criminalization framework is still in development.

Caveats

It is important to understand what this vote does not do.

First, this is a vote by the European Parliament. The final text must go through the full legislative process, including the Council of the EU, before becoming binding law. The current status is a reported vote on a revision.

Second, the ban targets the tools. It does not automatically delete every existing deepfake image already circulating on the internet. Removal of existing content often relies on separate mechanisms, such as reporting platforms.

Third, the effectiveness depends on enforcement. Banning a tool is different from preventing its use. The law must define penalties for developers and users who bypass these restrictions.

Finally, the source material notes that the current measures might be considered "too light" by some observers. The ongoing project to criminalize the act suggests that the tool ban is part of a broader, evolving strategy, not a final solution.

What’s next

The immediate next step involves the finalization of the criminalization bill. The European Union is currently drafting legislation to make the creation and distribution of these deepfakes a criminal offense.

Simultaneously, industry players are building infrastructure to help victims. A key tool in this ecosystem is StopNCII. This is a free, global reporting platform that allows victims of revenge porn and deepfakes to submit their original photos.

The goal of StopNCII is to track and remove non-consensual content from the web. It operates by leveraging a network of major web actors. Platforms such as Meta (including Facebook, Instagram, and Threads), Snapchat, TikTok, Reddit, OnlyFans, and PornHub already participate in its development.

Microsoft has also joined this effort. The company announced a partnership with SWGfL, the non-profit administering StopNCII, to detect and remove illicit content on its Bing search engine.

For victims, the path forward involves using these reporting tools while the legal framework solidifies. The combination of a tool ban and a criminal bill aims to close the loop: preventing creation and punishing distribution.

Going further

• 20 Minutes: How the EU is tackling pornographic deepfakes — The original report detailing the European Parliament vote and the scale of the Grok incident.
• Clubic: Microsoft joins the fight against revenge porn on Bing — Details on the StopNCII platform and industry partnerships for content removal.
• EUR-Lex: The AI Act — The official text of the EU AI Act adopted in 2024, which serves as the basis for these revisions.
• StopNCII: Reporting platform — The official portal for victims to report and request the removal of non-consensual intimate images.