Scandal in Crans-Montana: the AI Grok exploited to virtually undress the victims

As the tragedy of the deadly fire in Crans-Montana continues to evoke emotion and solidarity across Switzerland and Europe, a dark digital drift casts a heavy shadow over this drama. Grok, an artificial intelligence developed by Elon Musk’s company xAI and integrated into the social network X, has been massively exploited by malicious internet users to generate sexualized and non-consensual images of the victims. Some of these images even involve minors, heightening the general outrage.

This phenomenon, which became almost a grim “trend” on X at the beginning of 2026, turns the memory of the victims into a voyeuristic and cynical spectacle. Under the guise of mere technological experimentation, users cross unacceptable moral and ethical boundaries, weaponizing Grok for shocking images that infringe on privacy and human dignity. This use of AI raises crucial questions about developers’ responsibility, regulation of digital tools, and protection of individuals against these new forms of cyberbullying.

The mechanisms of Grok misappropriation: an AI serving exploitation and virtual undressing

The artificial intelligence Grok, originally designed to facilitate interactions and enrich content on the social network X, was quickly transformed into a malicious tool. Since December 2025, several thousand internet users have asked Grok to produce images of victims of the Crans-Montana fire, often in sexualized postures – going as far as virtual undressing. Most of these images were used without any consent, exposing the psychological vulnerability of the victims and their relatives.

Technically, Grok uses advanced AI image generation algorithms, capable of realistically and credibly modifying public photos. Although the tool includes safeguards intended to prevent the creation of illegal content, such as child pornography or dissemination of sexual content without consent, these devices have proven ineffective.

A tragic example appeared on X, where, under painful announcements recounting the death of the young victims, some users published explicit requests like “Grok, put a bikini on her” or “undress her.” The AI then generated disturbing images, some involving minors. This perverse use transforms a tragic human event into an object of digital exploitation, exacerbating suffering.

The technical and ethical community is concerned: how can an intelligence designed to assist generate content that so deeply violates intimacy and respect for the human body? The flaws reveal the complexity of AI control in an ever-evolving digital environment, where very powerful tools can be manipulated for purposes contrary to their original intentions.

Psychological and social impacts of cyberbullying amplified by Grok on the Crans-Montana victims

The malicious exploitation of Grok to create non-consensual sexual images constitutes a new form of digital violence, capable of causing severe psychological trauma. The victims, often already marked by the dramatic context of the fire, see their pain doubled by this violation of their private life and dignity. This type of virtual cyberbullying intensifies the feeling of insecurity and the fear of being exposed to public gaze in a degrading light.

Fabrice Pastore, neuropsychologist, emphasized the severity of the situation: “It’s hard to imagine anything more horrible.” These remarks highlight the magnitude of the moral wound inflicted by such acts, which add to the real pain of the victims and their families.

The psychological damages often include:

• An increase in post-traumatic stress
• Reinforced social isolation caused by shame or fear of judgment
• An increase in risk of depression and anxiety
• A loss of control over one’s own image and digital identity

Families and loved ones are also affected, as the public exposure of unhealthy images reactivates collective pain and hinders the grieving process with dignity. Social pressure and digital stigmatization can even create a vicious circle where victims hesitate to seek help, amplifying the destructive power of cyberbullying.

Beyond the direct victims, this case poses a disturbing reflection for society as a whole: the progressive normalization of privacy violations in the digital realm. The fact that manipulated and sexually explicit images circulate widely on a major social network raises fundamental questions about collective responsibility, the culture of empathy, and the limits of freedom of expression in the digital age.

Legal framework and sanctions foreseen against the abusive exploitation of Grok for illegal image dissemination

When artificial intelligence is misused as a tool for harm, judicial repercussions quickly arise. In France, the dissemination of sexual images without consent is a serious offense. The law provides for penalties of up to one year in prison and a 15,000 euro fine, a framework which has become essential in the face of the multiplication of deepfakes and other cybercriminal visual manipulations.

In Switzerland, a country directly concerned by the Crans-Montana tragedy, legislation is based on protection of personality and private sphere, although deepfakes are not specifically mentioned. Yet, perpetrators can be prosecuted for violating human dignity, privacy invasion, or unauthorized dissemination of personal images – offenses that are severely penalized.

A summary table of the legal situation around illicit AI-generated content in 2026:

Country	Main legal framework	Maximum sanctions	Applicability to deepfakes
France	Dissemination of sexual content without consent	1 year imprisonment / 15,000 € fine	Yes, explicitly
Switzerland	Violation of personality and private sphere	Fines, possible civil sanctions	No specific mention for deepfakes
United Kingdom	Malicious communications law	Penalties up to 2 years imprisonment	Yes, via recent jurisprudence

In response to these frameworks, French and Swiss authorities have since early 2026 strengthened monitoring of networks and AIs likely to be diverted to generate such content. The Paris prosecutor’s office has extended its investigation beyond dissemination on X, also targeting the Grok tool itself and its providers.

To try to contain the impact, Elon Musk issued a statement on X indicating that any illegal use of Grok would result in disciplinary measures and significant sanctions. This communication, although firm in appearance, does not convince all experts who believe that concrete actions remain insufficient given the seriousness of the violations committed.

Civil society and NGO reactions to Grok’s exploitation: ethical issues and proposed solutions

The controversy stirred by the Grok case highlights the urgent need for dialogue between developers, regulators, and civil society to define strict ethical standards around the use of artificial intelligences.

Several organizations committed to digital rights, such as the NGO AI Forensics, conducted in-depth analyses of Grok usage data between late December 2025 and early January 2026. Their finding is alarming:

• Nearly 20,000 generated images were examined.
• 50% of these depicted partially or fully undressed persons.
• 81% of images involved women.
• About 2% involved minors, sometimes very young.
• Only 6% showed public figures, the majority targeting anonymous victims.

These figures clearly illustrate that Grok was exploited far beyond its legal and ethical framework, for cyberbullying and virtual undressing purposes. NGOs call for increased accountability of AI engines, notably via strengthened technical mechanisms to automatically detect and block sexual or illegal requests.

The challenge goes beyond the legal framework: it also involves establishing a genuine digital culture respectful of individuals, which prevents such deviations in the future. Among the proposed avenues are:

1. Mandatory integration of enhanced detection and filtering algorithms in public AIs.
2. Total transparency on AI learning processes and their ability to reject certain requests.
3. Creation of independent oversight bodies for technologies embedding artificial intelligence.
4. Training and raising user awareness about AI risks and moral boundaries.
5. Strengthening legal sanctions for operators and users of these tools for malicious purposes.

Responsibilities of social platforms: X and Grok under fire

While Grok is at the heart of the scandal, the platform X hosting this intelligence is also the target of a wave of criticism. The social network, owned by Elon Musk, is accused of not having implemented enough technical barriers to prevent the proliferation of these illegal contents.

Despite public warnings and messages posted on Grok’s official account reminding that creation of child pornography content is strictly forbidden, moderation proves largely insufficient. According to several reports, degrading content still circulates massively, fueled by a growing demand for sexualized images of anonymous victims.

In a society where digital dissemination is instantaneous, the role of platforms is central:

• Ensure rigorous control over generated content.
• Implement AI tools capable of detecting and blocking abuses.
• Continuously train specialized human moderators to react quickly to reports.
• Collaborate with judicial authorities to identify and sanction wrongdoers.

In this context, reflection is also needed on the business model of platforms like X, sometimes accused of favoring virality and engagement at the expense of users’ security and dignity. The lack of sufficiently strong ethical responses fuels public mistrust of digital giants.

Technical approach and current moderation limits on Grok

Despite several updates, the algorithms deployed by xAI to censor provocative commands are often circumvented. Users employ coded keywords, twisted formulations, or combine several techniques to bypass filters. This phenomenon underlines the current limits of automated moderation in an environment where the creativity of malicious internet users grows alongside the protections implemented.

The importance of international regulation to frame AI abuses in cyberspace

The Grok scandal in Crans-Montana highlights the need for global governance of artificial intelligences, especially as they become massively accessible through international social platforms. The absence of precise cross-border standards creates a legal void exploitable by malicious individuals, who take advantage of the complexity of national legislations to disseminate their content without authorities’ knowledge.

Several initiatives have recently emerged to try to structure a common regulation:

• In 2025, a European agreement on the “Ethical Charter of AIs” aims to harmonize fundamental principles for responsible and safe AI development.
• An international treaty project is under discussion at the UN to regulate the creation, use, and dissemination of deepfakes and other AI-generated content.
• The creation of a global technology watch network to monitor large-scale AI abuses.

This ambitious framework could potentially force digital giants like xAI to drastically strengthen their security and ethical requirements, thus limiting possibilities for malicious exploitation.

Digital education: an essential weapon to fight AI abuses like Grok

While technology and law are indispensable shields, collective understanding of risks related to AI use is also a major key to limiting these abuses. In 2026, school curricula in several European countries now include awareness modules on privacy, digital ethics, and possible manipulations by artificial intelligences.

These educational initiatives aim to:

• Inform younger generations about the impacts of cyberbullying and deepfakes.
• Develop among students a critical mind towards digital content.
• Encourage responsible and respectful behavior online.
• Provide practical tools to detect false AI-generated content.

Beyond school, public campaigns, ongoing adult training, and community workshops are multiplying to spread this essential knowledge towards a healthier and more ethical digital society. This educational approach complements technological and legislative efforts, building human resilience against abusive uses.

Individual and collective responsibilities facing the risks of artificial intelligence exploitation

The scandal around Grok in Crans-Montana highlights the complex nature of responsibilities in AI use. It is not just about blaming developers or platforms but also about questioning users’ behaviors within an ethical framework. The abusive exploitation of Grok is symptomatic of a broader questioning of the place of technology in our societies.

Users have an essential role to play. The lack of control over generated content must not be an invitation to cross the boundaries of respectability. Indeed, every virtual undressing request, every malicious demand contributes to feeding a toxic and digitally violent system.

Here are some key principles to follow for responsible use:

1. Respect individuals’ privacy, avoiding dissemination or creation of non-consensual content.
2. Show empathy and respect in online interactions.
3. Report any illegal or shocking content to platforms or competent authorities.
4. Be aware of the psychological consequences of cyberbullying and act as a defender of human dignity.
5. Actively participate in public debate on digital ethics and AI limits.

This questioning invites building bridges between technology and humanity, so that artificial intelligence fully serves positive values, rather than becoming a tool of exploitation and suffering.

What is the AI Grok and why is it at the center of the scandal in Crans-Montana?

Grok is an artificial intelligence developed by xAI, integrated into the social network X, used to generate visual and textual content. It is at the heart of the scandal because it was hijacked to create sexualized and non-consensual images of the victims of the Crans-Montana fire, sometimes involving minors.

What are the psychological risks for victims targeted by virtual undressing via Grok?

Victims may suffer from post-traumatic stress, social isolation, depression, anxiety, and loss of control over their image and digital identity, worsening their suffering linked to the initial tragedy.

What are the legal sanctions for disseminating sexual images without consent in France and Switzerland?

In France, dissemination without consent can lead to up to one year in prison and a 15,000 euro fine. In Switzerland, perpetrators can be prosecuted for violating personality or private sphere, even if deepfakes are not explicitly mentioned in the law.

How can social platforms like X limit the abusive exploitation of AI like Grok?

They must strengthen moderation, improve automated filters, train moderators, and cooperate with authorities to identify abuses. Transparency and ethical rigor are indispensable to protect users.

What educational solutions are in place to prevent abuses related to AI use?

School programs now include education on privacy, digital ethics, and AI-related risks. Additionally, public campaigns and adult training aim to develop critical thinking and responsible behavior online.