One in Six MEPs Targeted by Sexualised AI Deepfakes, Study Finds

A policy brief published by the think tank Agora Digitale Transformation in October 2026 reports an audit documenting 50 instances of sexualised AI‑driven deepfakes targeting Members of the European Parliament (MEPs). According to the document, 46 of the affected MEPs are women. The study was publicly presented and its findings were highlighted by the Greens/EFA group in the European Parliament; several European news outlets subsequently covered the results.

The report uses the term 'sexualised deepfakes' to describe imagery or audiovisual manipulations that portray people in pornographic or sexually degrading contexts without their consent. Rather than claiming to be a comprehensive register of every instance, Agora describes its publication as an audit of publicly accessible examples, identifying patterns in dissemination, visibility and political impact, and situating these cases within the existing legal and institutional landscape.

The cases documented by Agora span a range of manipulations: from doctored photographs and misattributed images to video‑like deepfakes that reproduce facial movements and voice. The technical sophistication varies – some examples are crude montages while others leverage contemporary generative models. The report highlights that the social and political effect of the content often matters more than its technical quality: even less polished fakes can cause significant reputational damage if propagated widely.

According to the brief, the manipulated materials appeared across diverse platforms and distribution channels, including social networks, closed messaging services and specialised websites. Agora points out that distribution pathways evolve rapidly, and that automated amplification mechanisms such as bot networks or algorithmic recommendation systems can exponentially increase the reach of even low‑quality forgeries. The study also notes limitations in forensic detection: while detection algorithms can provide evidence, manipulators continually adapt methods, complicating attribution and removal.

The Greens/EFA group in the European Parliament picked up the report and has called for stronger protections for parliamentarians, with a particular emphasis on women who are disproportionately affected. In a press release, the group used the study's findings as a basis to press for faster intervention by platforms and more comprehensive support for victims of digital violence.

Beyond the immediate political reactions, civil society actors have framed the issue as a threat to democratic participation. If elected representatives face targeted digital abuse, including sexualised deepfakes, this can impair their capacity to perform public duties and deter others from seeking public office. The study has therefore intensified debate about prevention measures, reporting mechanisms and the resources needed to support affected MEPs.

Agora's analysis suggests that current legal frameworks and platform obligations do not always provide adequate protection. The European Union has enacted laws in recent years that touch on online intermediaries and AI systems, including the Digital Services Act (DSA) and the AI Act. However, the report argues that implementation gaps and operational challenges can hinder effective protection against sexualised digital violence.

The brief and subsequent political statements emphasise practical obstacles: reporting and redress mechanisms are often slow or fragmented, especially when content spreads across multiple jurisdictions and services. Questions of liability also remain contentious — for instance, how platforms should balance content moderation with due process, who bears the cost of forensic analysis, and how cross‑border cooperation should be organised in investigations.

The Agora report and reactions from lawmakers underline a twofold effect. First, individual MEPs suffer direct harms: reputational damage, psychological stress and increased security needs. Second, there is concern about chilling effects on democratic engagement. Women who face a higher burden of targeted sexualised abuse may be discouraged from visible public roles or political participation.

Experts and advocacy groups stress that treating incidents solely as isolated scandals misses the structural dimension: gendered digital violence against public figures interacts with broader dynamics of online harassment, information disorder and political polarisation. The resulting environment can narrow public debate if targeted individuals withdraw or alter their behaviour to avoid abuse.

Despite clear counts in the Agora audit, key aspects remain unresolved. Attribution of deepfakes — identifying authors or orchestrators — is technically and legally complex, and the study does not name perpetrators in most instances. It is also unclear to what extent organised campaigns versus individual actors account for the documented cases.

Agora and supportive MEPs call for a range of measures: faster take‑down and cross‑platform coordination, specialised support services for victims, and improved forensic and reporting capabilities. The brief stresses that technological detection alone will not suffice; legal clarity, resource allocation and institutional cooperation are necessary. How EU institutions, national authorities and platforms will translate these recommendations into concrete policy or enforcement action will be a focal point of debate in the months ahead.