German actress Collien Fernandes is the victim of sexual deepfakes created and shared to other men by her husband. In Germany as elsewhere in Europe, this scandal exposes the flaws in criminal law that is still poorly equipped to deal with generative AI, while 98% of deepfakes are pornographic and almost exclusively target women.
Published on: 04/16/2026 - 09:25
“My body was stolen.” For around ten years, fake, rigged pornographic images featuring German actress and presenter Collien Fernandes were created and distributed on the Web, giving the illusion of private content shared by accounts in her name.
At the end of 2024, the actress, who had been searching for the authors of these montages for years, discovered that the person behind these deepfakes was none other than her husband: Christian Ulmen, also an actor and from whom she is now divorced. Facts that he himself admitted last March after she filed a complaint, admitting to having created false profiles and evoking, in an exchange with lawyers, an uncontrollable "sexual fetish", reports Der Spiegel.
In Germany, the Collien Fernandes affair is making a lot of noise. The actress's lawyer even mentions a "digital Pelicot affair", in order to encourage her client to file a complaint.
Also read “Historical trial” of the Mazan rapes: after the shock wave, the legacy in question
“That sentence really stuck in my head and I told myself that I couldn’t let him get away with that,” Collien Fernandes told Agence France-Presse (AFP).
Behind this case, deepfakes are emerging as a new form of sexual violence, revealing the limits of the European legal framework in the face of these uses.
“Virtual rape”
The 44-year-old actress today accuses her ex-husband, the actor and host Christian Ulmen, 50, of having made around thirty men, including some from her professional entourage, believe that they were maintaining a secret online relationship with her, sometimes for years, by sending them false nude images or by having telephone conversations of a sexual nature by doctoring her voice.
During a dinner with a producer, she learns that he thought he had exchanged sexual messages with her on the Web. Then, she discovers a fantasy scenario that her husband wrote and broadcast, in which she is raped by 21 men.
“You virtually raped me,” she accuses, addressing her ex-husband. After the latter's confession, Collien Fernandes immerses himself in the Mazan rape case. An “incredibly hard” affair, which resonates within her to the point that she cannot read an entire article without “bursting into tears”, she told AFP, on the verge of tears.
It is “the strength” that Gisèle Pelicot displays that gives her the courage to file a complaint, to put an end to rumors about her and take up the fight for victims of digital violence.
To (re)readThe rise of masculinism (1/3): digital violence and hatred of women
“Sexual deepfakes are part of a relationship of gender domination and a continuum of sexist and sexual (cyber) violence, which mainly affects women and LGBTQIA+ people,” recalls the Hubertine-Auclert Center.
Worse: 98% of deepfakes produced are pornographic and 99% of people targeted are women, according to the “State of Deepfakes” research conducted in 2023 by the start-up fighting online identity theft Security Hero.
The democratization of generative artificial intelligence (AI) tools makes the creation of deepfakes very easy, increasing the risks for women and girls. “Hyper-faking featuring one’s privacy can create a feeling of helplessness, loss of control over one’s image, dispossession of oneself,” underlined Célia Zolynski, professor of private law and criminal sciences, in the podcast “Les Voies de l’AI” in February 2025.
In Belgium, the Institute for the Equality of Women and Men (IEFH) deplores in a press release that AI acts as an “accelerator” of digital violence against women.
In the United States, according to a recent study, more than half of deepfake victims have considered suicide. “This is not a marginal Internet problem. This is a global crisis,” says UN Women.
Legal gaps and discrepancies in law enforcement
“Digital violence is REAL violence!”, alarms Renew Europe on X, noting the “alarming legal flaws” in the Collien Fernandes affair. “We demand a ban on 'nudification' AI apps and stronger protection for victims of deepfakes in the EU,” continues the European parliamentary group (center).
In a press release published in the wake of the affair, the political group calls for “reinforced European action, in order to fill the legal gaps and the divergences in the application of laws allowing the abusive use of artificial intelligence to generate intimate non-consensual content”.
Collien Fernandes' complaint was filed in Spain, where the ex-spouses reside and where the law regarding online violence is stricter. But on Tuesday April 14, the complaint was finally closed, the court considering that it did not have jurisdiction and thus transmitting the procedure to the German public prosecutor's office.
On this subject, Germany is currently lagging behind. The creation of a deepfake is not, as such, clearly captured by the existing incriminations. In a report, the Bundestag explicitly says that non-consensual pornographic deepfakes of adults are “taken into account in a limited manner”.
For comparison, France, which is among the most advanced countries in terms of criminal classification, adopted in 2024 the law on the regulation of the digital space, or SREN law, explicitly targeting sexual content generated by algorithms and reproducing the image or voice of a person without their consent.
In other European countries, including Germany, the transposition of EU Directive 2024/1385 will fill this gap. This directive is in fact the most important common framework on violence against women. It requires member states to criminalize the non-consensual sharing of intimate images, including deepfakes, and the manipulation or creation of sexually explicit material without consent. States have until June 14, 2027 to transpose it.
But the Collien Fernandes affair could move the lines faster than expected across the Rhine. “This subject is of capital importance, which is why we began drafting a bill within the ministry very early on. This project is now finalized. In the future, we want to criminalize the creation and distribution of pornographic deepfakes,” announced the German Federal Minister of Justice, Stefanie Hubig, reacting to the Fernandes affair on March 20 in a video.
“We want to regulate criminal liability linked to deepfakes in general. We also want to facilitate victims' access to justice, enable them to obtain compensation more quickly and thus put an end to the humiliation and degradation to which they are victims online.”
To display this Instagram content, it is necessary to authorize audience measurement and advertising cookies.
These acts will be criminalized by a law on digital violence which will be presented "in a very short time", for his part specified the spokesperson for the German Ministry of Justice, Eike Hosemann, conceding that "criminal law is not up to what it should be, in particular when it comes to deepfakes".
On March 26, the European Parliament approved a measure to ban services that use AI to “create or manipulate sexually explicit or intimate images.”
This initiative notably followed the introduction a few months ago of a feature in Grok – Elon Musk's AI assistant – which allowed users to ask him to create sexual deepfakes of naked women and children from real photos.
Read alsoGrok, Elon Musk's AI, caught up in the sexual deepfakes controversy
For Collien Fernandes as for others, the violence did not stop at the private sphere. It has continued, transformed, diffused, in a space where the law still struggles to contain it.
Keywords associated with the article
The content you are trying to access does not exist or is no longer available.




