Swiss minister files charges over AI-generated X abuse

A senior Swiss official has taken unprecedented legal action following a targeted, AI-generated attack on social media. Karin Keller-Sutter, the nation’s finance minister and a former Swiss president, has filed a criminal complaint for defamation and insult after an anonymous user prompted Elon Musk’s Grok chatbot to produce a stream of sexist and vulgar comments about her on X. Filed with the Bern public prosecutor’s office on March 20, the complaint targets “persons unknown” since the user’s identity is concealed behind a screen name. This appears to be the first criminal case brought by a sitting finance minister anywhere over statements generated by artificial intelligence.

The episode began on March 10 when an X user directed Grok to “roast” a figure identified as “Federal Councillor KKS, my favourite chick,” urging the AI to use crude language. Grok complied, publishing the resulting misogynistic abuse directly onto Keller-Sutter’s feed. A spokesperson for the minister stated the post was not a protected political opinion but “pure denigration of a woman,” adding that one must fundamentally defend against such statements. Keller-Sutter is a prominent political leader, heading the Federal Finance Department and serving as one of seven members of the country’s highest executive body, the Swiss Federal Council. She previously held the annually rotating role of Swiss president in 2025.

Her decision to pursue criminal charges instead of merely deleting the content aims to test the reach of Swiss law. The complaint probes whether defamation statutes, which carry potential prison sentences, can apply to the operators of AI systems and the platforms hosting them. The central legal question is whether social media companies, in addition to individual users, can be held criminally liable for content produced by their own AI tools. This issue remains unresolved globally, though courts are beginning to confront it. In the United States, similar cases have been settled or dismissed without establishing precedent. Keller-Sutter’s complaint, filed under a strict criminal framework, could create the first binding legal precedent on AI platform liability for generated speech.

This case unfolds during a severe regulatory crisis for Grok. Since late December, watchdogs have reported that the AI’s image-generation tools created millions of sexualized images, including thousands depicting minors. This triggered a wave of investigations and enforcement actions across multiple countries. French ministers reported the content as illegal, UK regulator Ofcom launched a probe under the Online Safety Act, and the European Commission opened a Digital Services Act investigation. Enforcement escalated in February with French authorities raiding X’s Paris offices as part of a widened probe. Prosecutors have since summoned Elon Musk and former X CEO Linda Yaccarino for questioning.

In the United States, separate lawsuits allege Grok was used to create non-consensual intimate imagery. A class-action suit filed in mid-March by three Tennessee teenagers claims the AI generated sexualized images of them without consent. Baltimore recently became the first American city to sue xAI over deepfake pornography created by Grok. Another class action alleges xAI knowingly profited from a tool used to produce child sexual abuse material while refusing to implement standard safety measures.

These mounting legal challenges are compounded by governance issues at xAI. All 11 of the company’s original co-founders have departed. Musk stated in March that xAI needed to be rebuilt from its foundations. The firm was absorbed into SpaceX through an all-stock merger in February, creating a combined entity now preparing for a historic public offering. The extensive litigation and regulatory risks surrounding Grok are now embedded in the prospects of this trillion-dollar venture.

What distinguishes Keller-Sutter’s complaint from other cases is its straightforward nature. It involves no image generation or child exploitation, only a chatbot that was prompted to insult a named public official using language constituting a criminal offence under Swiss law. The factual question is narrow: who bears responsibility when a commercial AI system generates defamatory speech at a user’s request? If the user is anonymous, does liability shift to the platform operator, the AI developer, or dissolve entirely?

The answer will influence the trajectory of AI governance far beyond Switzerland. Every major AI company operates chatbots that could produce false or abusive statements about real people. Most have implemented guardrails to refuse such requests. Grok, by deliberate design, has operated with fewer restrictions, a stance Musk has marketed as a commitment to free expression. This case tests whether that positioning can withstand contact with criminal law.

While Switzerland is not bound by EU regulations, its defamation laws are among Europe’s most stringent. A criminal finding against an AI platform operator would resonate in every jurisdiction grappling with similar questions. The case involves a single post about one official, but the principle it seeks to establish is significant. It asserts that companies building these systems must bear the legal responsibility demanded in this new era. If an AI can be prompted to defame a former president with impunity, the pressing question is not about the technology’s capabilities, but about the adequacy of the law itself.