The Weaponization of Likeness: How European Politicians Are Being Targeted by an Industrialized Deepfake Pornography Ecosystem

Nearly 150 elected officials across Europe have been identified as targets within a sprawling, industrial-scale network of deepfake pornography websites, according to comprehensive new research that highlights a disturbing intersection of artificial intelligence and political intimidation. The study, conducted by Benjamin Shultz, a fellow at Agora Digitale Transformation and lead researcher at the American Sunlight Project, reveals that female members of parliament (MPs) are disproportionately targeted, facing a risk 33 times greater than their male counterparts. This data underscores a widening crisis in digital security, where the ease of generating nonconsensual synthetic media is being leveraged to harass and undermine women in public office.

The Scope of the Digital Threat

The investigation, which scrutinized approximately 160 domains known for hosting nonconsensual sexual content, focused on national parliaments across all 27 European Union member states. The findings are sobering: at least 138 female MPs have appeared in deepfake imagery or have been listed in specialized database entries designed to facilitate the creation of such content. In contrast, only nine male MPs were found to be referenced within the same ecosystems.

The digital abuse manifests in several forms. Some sites host fully realized explicit videos featuring the likenesses of political figures, while others maintain "profiles" that aggregate official photographs, social media handles, and personal biographies. These profiles often function as gateways, pairing the victims’ identities with sophisticated, automated "nudify" applications. Within just a few clicks, a user can transition from a professional portrait of a legislator to an AI-generated, sexually explicit image. This architecture of abuse transforms these websites into functional, profit-driven businesses, where the barrier to entry for generating harassment is effectively eliminated.

A Chronology of Escalation

The emergence of deepfake technology as a tool for harassment traces back to late 2017, when the first AI-generated videos began appearing online, largely targeting celebrities. The evolution from high-effort, niche experimentation to an automated, mainstream service has been rapid.

• 2017–2018: Early deepfake tools emerge, requiring significant technical expertise. Initial victims are primarily high-profile celebrities and public figures.
• 2019–2021: The proliferation of "nudify" apps gains momentum. The technology becomes increasingly accessible to the general public, moving from sophisticated software to consumer-facing, cloud-based interfaces.
• 2022: Evidence emerges of deepfakes being used in educational settings, with reports of school-aged children being targeted by peers.
• 2023–2024: The focus of these networks shifts toward political figures. Researchers begin mapping the scale of the abuse, identifying that these sites are not merely hosting content but actively inviting users to generate new, custom material of specific, identifiable women.
• 2025: Current research confirms the systematic targeting of European legislators, marking a transition from random harassment to the potential weaponization of AI against democratic institutions.

Data-Driven Analysis of the Ecosystem

The research highlights a clear correlation between the seniority of an official and the frequency with which they are targeted. Politicians from Germany, the Netherlands, Italy, and France were found to have the highest representation on these platforms, suggesting that the perpetrators may be seeking to maximize the perceived "impact" or "shock value" of the imagery by targeting high-profile figures.

The "industrialization" of this phenomenon is perhaps the most concerning aspect. Henry Ajder, a prominent AI advisor who has monitored the synthetic media landscape since 2018, notes that these platforms operate with the efficiency of legitimate tech companies. The infrastructure is designed to scale; it relies on automated search engine indexing and user-contributed content to ensure that if one site is shuttered, the content remains available elsewhere. The data shows that these sites are not just repositories—they are engines of production that monetize the violation of privacy.

Official Responses and the Quest for Remedy

The revelation has sent shockwaves through European political circles. Suzanne Kröger, a Dutch member of parliament, described the findings as "deeply disturbing" and representative of a broader, systemic trend of online intimidation against women. For Kröger, the issue is not merely personal; it is an attack on the integrity of the political process. "We are politicians, we have a platform and speak out," she stated. "But meanwhile, there are millions of women victims of these criminal acts."

Similarly, Sarah Dobbe, an MP for the Dutch Socialist Party, emphasized the urgent need for a robust support framework. The challenge, according to Dobbe, is twofold: providing immediate psychological and legal support to victims while ensuring that the infrastructure—the websites and the software—is dismantled with greater speed.

In response to these findings, researchers have begun the arduous process of notifying the affected MPs, providing them with technical guidance on how to report content to host platforms and, where possible, initiate takedown requests. However, this remains a game of digital "whack-a-mole," where the volume of content often outpaces the capacity of legal or platform-based moderation systems.

Broader Implications for Democracy

The chilling effect of such technology on political participation cannot be overstated. When the cost of entering public life includes the threat of perpetual, nonconsensual sexual harassment, the pool of potential candidates is inevitably narrowed.

"This discourages women from taking office in the first place and also crowds these women out of public life," Shultz notes. The psychological toll of being the target of a deepfake campaign is significant, often resulting in prolonged trauma, damaged reputations, and, in some cases, the premature abandonment of professional ambitions.

Furthermore, the implications for democratic health are profound. If the public perceives that their representatives are being systematically undermined through synthetic media, the trust in information—and by extension, the trust in democratic processes—begins to erode. The ability to distinguish between genuine political discourse and manufactured, abusive content is becoming a requisite skill for both the public and the political class.

Legislative Hurdles and the Path Forward

While legislative bodies have historically lagged behind the rapid pace of AI innovation, there is a perceptible shift in the regulatory landscape. Both the United Kingdom and the European Union have signaled intentions to implement strict bans on "nudify" services, aiming to criminalize the underlying business models of these sites rather than just the individual pieces of content.

In the United States, the "Take It Down Act" represents a significant step toward addressing the legal void, having already contributed to the forced closure of several major platforms that hosted thousands of images of celebrities and public figures. These legal precedents are being watched closely by European policymakers, who are under increasing pressure to synchronize their efforts.

The success of these interventions is measurable. Shultz’s previous analysis of US senators and members of Congress targeted in 2024 showed a "significant decrease" in depictions following the removal of key websites. This suggests that while total eradication may be a distant goal, targeted enforcement and the dismantling of distribution hubs can drastically reduce the prevalence of this content.

Conclusion: A Call for Sustained Action

The research into European MPs provides a critical snapshot of a rapidly evolving crisis. It is no longer an abstract fear of "future technology" but a present-day reality that threatens the safety and dignity of public servants. As the AI ecosystem continues to mature, the necessity for a multi-layered approach—combining aggressive legislative action, improved platform accountability, and increased public awareness—is paramount.

For the women currently navigating this landscape, the path forward remains fraught. As Kröger noted, the priority must remain two-fold: ensuring the rapid removal of abusive content and establishing a global regulatory standard that treats the nonconsensual generation of synthetic imagery as the serious criminal act that it is. The survival of an inclusive and robust democracy may well depend on the collective ability to protect its participants from the shadow of the deepfake.