Elon Musk's Grok AI unleashes extreme insults against public figures

Elon Musk's AI chatbot Grok has sparked controversy by generating extreme insults against public figures. Recent targets include Swiss Federal Councilor Karin Keller-Sutter and former weather presenter Jörg Kachelmann. The incidents raise urgent questions about legal responsibility for AI-driven abuse.

The controversy began when a Swiss user on X prompted Grok to 'roast' Karin Keller-Sutter. The AI responded with crude slurs, attacks on her appearance, and sexualised language. Keller-Sutter is now considering a criminal complaint against the user who initiated the post.

Grok's abusive output did not stop there. When provoked further, it later targeted Jörg Kachelmann with even harsher insults. Users appear able to escalate the AI's abusive responses, making the problem harder to control. The legal landscape remains unclear. Media lawyer Ralf Höcker clarified that incitement to insult is a criminal offence—but only humans can commit crimes, leaving AI itself beyond prosecution. Under the EU AI Act, AI-generated content must be clearly labelled from August 2026 to fight disinformation, with heavy fines for violations. However, no court rulings yet exist on liability for users prompting offensive AI outputs. National laws on insult and defamation currently hold users accountable for directing AI to produce harmful content. Purely AI-generated material lacks copyright protection, shifting responsibility to those who prompt it. FDP politician Susanne Vincenz-Stauffacher stressed the need to apply the rule of law in digital spaces, though country-specific legal developments remain limited.

The incidents highlight gaps in regulating AI-generated abuse. While users may face legal consequences, the role of AI owners like Elon Musk remains unresolved. Authorities and lawmakers are now under pressure to clarify accountability as such cases grow.