Non-Consensual Intimate Media

Non-Consensual Intimate Media, often abbreviated as NCII, refers to nude images or sexual recordings of a person that are shared without their consent. The English term literally means “intimate media without consent.” The material can be genuine, for instance a photo from a past relationship that is later posted online. But it can also be entirely fabricated: software generates a convincing-looking nude image from an ordinary vacation photo. What matters for the term is not whether the image is real, but that the person depicted did not consent. In professional circles, this is therefore also referred to as image-based sexual abuse. The older term “revenge pornography” is considered misleading, since revenge is by no means always the motive.

Why AI has massively worsened the problem

In the past, a perpetrator needed a genuine image. They had to obtain it, steal it, or secretly record it. This severely limited the number of possible victims. Since programs capable of generating images have become available, a single portrait photo from a social network is enough. This means that virtually anyone with a face photo online somewhere is vulnerable.

Then there is the effort involved. A few years ago, a convincing fake required technical knowledge and a lot of computing time. Today there are apps that do this in under a minute, some of them for free. Studies by organizations such as the UK’s Internet Watch Foundation show that the number of reported AI-generated abuse images has increased manyfold within just a few years. Women and girls are particularly affected, and at schools it is often female classmates.

For those affected, the harm is real even if the image is fake. The image circulates in class chats, ends up in search engines, and can hardly ever be completely removed. Many victims report anxiety, changing schools, or withdrawing from the internet. This is precisely why lawmakers now largely treat genuine and fake recordings the same.

How such images are created and how they can be taken down

The fakes are created using image generators, i.e. programs that compute new images from a text prompt or a template image. So-called nudify apps specialize in replacing the clothing in a photo with a computed body. Reputable providers block such requests. However, models freely available online can be subsequently modified to remove these safeguards.

There is a technical trick against distribution called hashing. A short numerical code, a digital fingerprint, is calculated from an image. Platforms compare every newly uploaded image against a list of such codes and block matches. The image itself does not need to be shared for this to work. This is exactly how the service StopNCII operates, allowing victims to report their images directly from their own device.

It isn’t perfect. If someone crops the image or changes the colors, the fingerprint may no longer match. And platforms that don’t participate in the system continue to spread the material. That’s why some countries additionally rely on takedown deadlines: a platform must then process a report within a fixed time, in the US for example within 48 hours.

NCII in the news, in laws, and in schools

In the news, the term usually comes up in two contexts. First, in high-profile cases, such as when fake images of singer Taylor Swift were shared millions of times in 2024. Second, in legislation: in 2024 the European Union passed a directive against violence toward women that also covers deepfake nude images. In the US, the so-called Take It Down Act came into force in 2025.

In Germany, distributing such images has long been a criminal offense, for instance through the protection of a person’s personal image under the criminal code and through personality rights. What is newly being debated is whether the existing rules are sufficient for AI-generated fakes. For providers of image generators, this is an economic issue: anyone who fails to prevent abuse risks fines, app store bans, and the loss of advertisers.

Most commonly, though, the term comes up in everyday school life. Nudify apps get passed around there as a prank, even though it constitutes a criminal offense. When minors are involved, the result is legally classified as child sexual abuse material, regardless of the fact that it was computer-generated. Anyone affected should save screenshots as evidence, report the material, and reach out to adults or counseling services.