What Was Banned
In July 2026, the European Commission\'s Digital Omnibus on AI added a specific prohibition to the EU AI Act: AI systems that generate, distribute, or facilitate non-consensual intimate imagery — commonly called "nudification" or "deepfake pornography" — are now explicitly banned under Article 5.
This prohibition was not in the original AI Act. It was added in response to the growing problem of AI tools that can generate realistic nude images from clothed photographs, and the increasing use of these tools to create non-consensual intimate imagery — often targeting women, minors, and public figures.
The nudification ban is not just about technology. It is about dignity, consent, and the right to control your own image.
What Is AI Nudification?
AI nudification tools use deep learning to generate realistic nude images from photographs of clothed people. The technology works by:
- Image analysis — The AI analyzes the input photograph to understand body pose, lighting, and proportions
- Body reconstruction — The AI generates a realistic nude body that matches the person\'s proportions and pose
- Face transfer — The AI preserves the person\'s face and other identifying features
- Image synthesis — The AI combines the generated body with the original face to produce a realistic nude image
These tools are available as mobile apps, web services, and open-source software. Some are marketed explicitly for creating non-consensual intimate imagery; others are marketed as "artistic" tools but are primarily used for non-consensual purposes.
Scope of the Prohibition
The EU AI Act\'s nudification ban covers:
- Generation — Creating non-consensual intimate imagery using AI
- Distribution — Sharing, uploading, or transmitting such imagery
- Facilitation — Providing tools or services specifically designed to create such imagery
- Knowledge — Knowingly or having reasonable grounds to know that the imagery is non-consensual
The prohibition applies to both individuals and companies. A company that provides a nudification tool can be held liable for how its tool is used, even if it did not directly create the imagery.
Why This Matters
Scale of the Problem
AI-generated non-consensual intimate imagery has exploded in recent years. According to a 2025 report by the Internet Watch Foundation, AI-generated child sexual abuse material increased by 450% between 2023 and 2025. Non-consensual deepfake pornography targeting adults has also increased dramatically, with some estimates suggesting that 96% of deepfake videos are non-consensual pornography.
Psychological Impact
Victims of non-consensual intimate imagery experience significant psychological harm, including anxiety, depression, PTSD, and suicidal ideation. The knowledge that realistic nude images of yourself exist and are being distributed can be devastating, regardless of whether the images are "real."
Chilling Effect
The availability of nudification tools creates a chilling effect on women\'s participation in public life. If any woman can be "nudified" from a public photograph, the threat of non-consensual intimate imagery becomes a tool of harassment and control.
How It Will Be Enforced
Enforcement of the nudification ban relies on several mechanisms:
- Platform responsibility — Online platforms are required to remove non-consensual intimate imagery when notified, and to take proactive measures to prevent its distribution
- Developer liability — Companies that develop and distribute nudification tools can be held liable if their tools are used to create non-consensual imagery
- Fines — Violations of Article 5 carry fines up to €35 million or 7% of global annual turnover
- Cross-border cooperation — The EU AI Office coordinates enforcement across member states, preventing companies from evading regulation by operating in different countries
Technical Challenges
Enforcing the nudification ban faces several technical challenges:
- Detection — Identifying AI-generated non-consensual intimate imagery is technically difficult. Current detection tools are not always reliable, and the quality of AI-generated images continues to improve.
- Open-source tools — Open-source nudification tools cannot be "shut down" because the code is freely available. Enforcement must focus on distribution platforms and commercial services.
- Encryption — Non-consensual intimate imagery can be shared through encrypted messaging services, making detection and enforcement difficult.
- Speed — Once an image is shared online, it can spread globally within minutes. Enforcement must be rapid to be effective.
EU vs. Other Jurisdictions
The EU is not alone in addressing AI-generated non-consensual intimate imagery, but its approach is the most comprehensive:
- United States — Several states have passed laws criminalizing deepfake pornography, but there is no federal law. The US approach is fragmented and inconsistent.
- United Kingdom — The Online Safety Act includes provisions against non-consensual intimate imagery, but does not specifically address AI generation.
- Australia — Australia has criminalized the distribution of non-consensual intimate imagery, including AI-generated content.
- Japan — Japan has laws against non-consensual photography, but they do not clearly cover AI-generated images.
What Can You Do
- Report non-consensual intimate imagery — If you encounter AI-generated non-consensual intimate imagery, report it to the platform where it is hosted and to the relevant authorities.
- Support victims — If someone you know is a victim of non-consensual intimate imagery, provide support and help them access resources.
- Advocate for stronger laws — Contact your elected officials about the need for comprehensive legislation against AI-generated non-consensual intimate imagery.
- Be cautious with your images — Be aware that any photograph of you can potentially be used to generate non-consensual intimate imagery. Consider the privacy settings on your social media accounts.
Sources
- European Commission, "Digital Omnibus on AI — Article 5 Prohibitions" — Legislative text, July 2026
- Internet Watch Foundation, "AI-Generated Child Sexual Abuse Material" — 2025 report
- Cyber Civil Rights Initiative, "Non-Consensual Intimate Imagery" — Research and resources
