When an App “Undresses” You: Why Europe’s Push to Ban Nudifier Apps is Necessary – but Not Enough

When an app “undresses” you, the violation is not just the image. It is the sudden loss of control over how your likeness can be used, reshaped, and circulated without consent.

Consider how easily this can happen. A completely ordinary photo – taken at the beach, at a party, or even pulled from a public profile – can now be run through an AI tool, often referred to as a nudifier, that generates a sexualised version in a matter of seconds. The output may be fabricated, but it is still recognisably tied to a real person. That makes the harm deeply personal even when the image is technically “fake.”

What makes this form of abuse particularly concerning is the low barrier to carrying it out. Nudifier tools are designed to be fast, accessible, and require little to no technical expertise, allowing users to generate manipulated images within seconds. This makes the abuse easy to replicate and scale, and more difficult to contain.

As these risks have become more visible, regulators have begun to respond. In March 2026, the European Parliament backed a proposal to ban these nudifier systems under changes to the AI Act, although the proposal leaves room for systems that claim to have effective safety measures (European Parliament, 2026). Around the same time, an Amsterdam court issued a preliminary injunction against xAI and its chatbot Grok in the Netherlands, ordering the company not to generate or distribute non-consensual sexualised images and attaching a €100,000-per-day penalty for non-compliance (Van Campenhout et al., 2026). The court found the system’s safeguards insufficient and ruled that Grok could not be offered on X while in breach of these requirements. This ruling is significant because it highlights the limits of relying on company-defined “safety measures” without independent verification.

These developments signal a broader shift in how regulators are beginning to approach the problem. AI-generated sexual abuse is no longer being treated only as harmful content to be removed once it appears, but is increasingly understood as a digital governance issue involving system design and platform responsibility.

Thus, Europe is right to target nudifier tools but banning them alone will not be enough unless governance also addresses platform amplification, opacity, and the burden placed on victims.

“Fake” images cause real harm

One of the misleading assumptions about synthetic sexual imagery is that it is less serious because it is not “real.” If no camera captured the nude body, the harm can appear less tangible. But that logic misses the point.

As scholars McGlynn, Rackley and Houghton (2017) argue, image-based sexual abuse extends beyond non-consensual photographs and includes the creation and distribution of sexual imagery without consent, regardless of whether the image is authentic or fabricated. From this perspective, nudifier abuse clearly falls within this category. The harm lies in the violation of consent, dignity, privacy, and autonomy, often resulting in reputational damage, emotional distress, and a loss of control over one’s own body.

Furthermore, this is not only a matter of abuse but a fundamental issue of digital rights. Privacy is often understood as controlling who can see your information. In this context, however, it also concerns how that information is used beyond its original setting. Nudifier technologies disrupt this by enabling images to be repurposed into intimate content. When these boundaries are broken, individuals can lose control over how their likeness is used, including the possibility that ordinary images are turned into pornographic material. This also raises security concerns, as the ease of generating such content undermines safe participation online.

Importantly, these harms are not distributed evenly. A European Parliament briefing projected that around 8 million deepfakes would be shared in 2025, up from 500,000 in 2023, with pornographic material accounting for about 98% of deepfakes (European Parliament, 2025). Research also shows that non-consensual intimate deepfakes disproportionately target women and girls, reinforcing existing gendered inequalities (Kira, 2024; Noble, 2018). This scale and its disproportionate impact on women and girls make clear that the issue cannot be treated as marginal but demands serious attention and response.

Why banning the tools is still not enough

Even if Europe succeeds in restricting nudifier apps, the harm does not stop at the point of creation. It can still extend through circulation. A single AI-generated sexual image can be downloaded, reposted, edited, and redistributed across multiple platforms within minutes. By the time a victim reports one post, copies may already exist elsewhere and are difficult to remove.

This persistence is not simply the result of individual users. Platform systems play an active role in shaping how widely content circulates. As Just and Latzer (2017) argue, algorithms shape what becomes visible and relevant, rather than simply organising information. In practice, this means that content attracting attention is more likely to be surfaced and sustained, regardless of whether that attention is driven by harm, shock, or exploitation.

Platform design further reinforces this dynamic. Massanari (2017) shows how features such as ranking systems, weak moderation, and engagement-driven visibility can allow harmful content to gain traction and remain accessible over time.

In the case of nudifier abuse, this dynamic is especially concerning. Sexualised images are more likely to attract attention, increasing their visibility. While the initial violation may occur through an AI tool, the broader harm is shaped by platform systems that enable continued circulation and exposure. Restricting tools may limit the initial act, but if platforms continue to sustain this visibility, governance addresses only part of the problem. This dynamic is further reinforced by the fact that these systems are developed by profit-oriented companies, where incentives to maximise engagement can amplify the visibility of harmful content.

Opacity makes the harm harder to escape

The persistence of the harmful content is also difficult to challenge because it is largely invisible to those affected. For victims, one of the most frustrating parts of the abuse is not knowing exactly how it happened or why it keeps happening. For example, an image may reappear across different platforms without any clear indication of where it originated or how it continues to circulate.

Pasquale’s (2015) concept of the “black box society” captures this imbalance well. He argues that many digital systems operate in ways that are opaque to the people affected by them, meaning that key decisions about visibility, ranking, and removal are hidden from public scrutiny. As a result, companies hold far greater knowledge and control over these processes than users ever will.

This opacity intensifies harm. If victims cannot trace how content spreads or understand moderation decisions, they are left navigating systems they cannot see or challenge effectively, limiting both personal redress and the ability to hold platforms accountable through effective oversight.

What could better governance look like

Taken together, these challenges point to the limits of current approaches and the need for a more comprehensive response. If Europe wants to address this problem effectively, it has to regulate the full chain of harm, from the creation of content to its distribution, amplification, and persistence.

That includes restricting tools designed to generate non-consensual sexual imagery – a step the European Parliament has begun to take (European Parliament, 2026). But it also requires independent testing of safety claims. If companies say their systems cannot produce harmful outputs, those claims should be verifiable rather than accepted at face value.

Governance must also address how this content is distributed across platforms. Features such as sharing functions, downloads, reposting tools, and cross-platform circulation enable images to move quickly beyond their original context. This speed also presents a challenge. While content can be generated and circulated within seconds, regulatory and moderation responses often operate more slowly, creating a gap in which harmful material can spread before effective intervention occurs. Addressing this requires not only stricter rules, but faster ones. Platforms could be required to implement rapid detection and response systems, such as automated flagging, immediate visibility reduction, and time-bound removal obligations for high-risk content.

Just as importantly, governance must address how platforms amplify and prioritise harmful content. Systems such as recommender algorithms, visibility rankings, and reposting mechanisms do not simply distribute content, but actively shape what becomes more visible and prominent, yet they remain largely opaque to users and regulators. Governance should therefore not only focus on the presence of harmful content, but also on how platform systems shape its visibility. This could include requiring greater transparency around ranking systems and imposing limits on the promotion of flagged or high-risk material.

Ultimately, the goal should not simply be to remove individual posts, but to address the systems that allow this content to persist, circulate, and remain difficult to challenge. This also requires recognising that these systems are not neutral infrastructures but are shaped by economic incentives that prioritise engagement and growth, raising the question of whether governance can remain effective if it does not also confront the underlying orientation of platforms toward profit rather than public interest. This may require new legal frameworks that treat major platforms as public-interest actors, imposing duties to prioritise harm prevention over engagement and subjecting their core systems to independent oversight.

The broader lesson

The debate over nudifier apps reveals a broader issue in digital governance. Online harm is no longer produced in a single moment by a single actor. It emerges from the interaction between AI systems, platform infrastructures, and governance frameworks. That is why Europe’s recent response is both necessary and incomplete. It is necessary because some technologies are too clearly structured around abusive uses to be treated as ordinary innovation. But it is incomplete if it stops at the tool and ignores the systems that give that harm social power.

The real issue is not only that AI can generate harmful images. It is that digital environments can make those images visible and difficult to escape.

Until governance addresses that broader chain – creation, circulation, opacity, and redress – people will continue to carry the burden of harms they never consented to.

And that is exactly what good digital governance should prevent.

References

European Parliament. (2025, July 3). Children and deepfakes. European Parliamentary Research Service.https://www.europarl.europa.eu/thinktank/en/document/EPRS_BRI%282025%29775855

European Parliament. (2026, March 26). Artificial Intelligence Act: delayed application, ban on nudifier apps. https://www.europarl.europa.eu/news/en/press-room/20260323IPR38829/artificial-intelligence-act-delayed-application-ban-on-nudifier-apps

Just, N., & Latzer, M. (2017). Governance by algorithms: Reality construction by algorithmic selection on the internet. Media, Culture & Society, 39(2), 238–258. https://doi.org/10.1177/0163443716643157

Kira, B. (2024). When non-consensual intimate deepfakes go viral: The insufficiency of the UK Online Safety Act. Computer Law & Security Review, 54, 106024. https://doi.org/10.1016/j.clsr.2024.106024

Massanari, A. (2017). #Gamergate and The Fappening: How Reddit’s algorithm, governance, and culture support toxic technocultures. New Media & Society, 19(3), 329–346. https://doi.org/10.1177/1461444815608807

McGlynn, C., Rackley, E., & Houghton, R. (2017). Beyond “revenge porn”: The continuum of image-based sexual abuse. Feminist Legal Studies, 25(1), 25–46. https://doi.org/10.1007/s10691-017-9343-2

Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. New York University Press.

Pasquale, F. (2015). The black box society: The secret algorithms that control money and information. Harvard University Press.

Van Campenhout, C., Ersen, H., & Reese, C. (2026, March 26). Dutch court rules against Grok over AI-generated “undressing” images in rare legal rebuke. Reuters. https://www.reuters.com/business/autos-transportation/dutch-court-orders-xai-grok-not-create-distribute-nonconsensual-sex-images-2026-03-26/