Elon Musk’s AI chatbot Grok generates suggestive images of women and children upon request. The company has admitted to reviewing the function. At management level, responsibility is denied. Several countries have launched investigations. A commentary analysis.
New Grok function degrades users
- Since the beginning of January, countless sexualized images have been circulating on X (formerly Twitter). The background is a function of the AI chatbot Grok, which is now part of the platform. With it, users can edit photos, including transforming normal photos of women or children into AI images with suggestive poses or bikinis.
- Many media outlets took up the topic and triggered a wave of outrage – also because the management of the parent company xAI was talking head and shoulders above their necks. The company spoke to the Reuters news agency about “lies from the established media”. However, after reality could no longer be denied, the function was admitted to being checked, but the criticism was dismissed.
- Elon Musk uses his Platform X and the AI chatbot Grok to achieve his own political goals. For example, he changed the algorithms in favor of numerous Trump supporters, specifically supports right-wing populist parties in Europe and rails against EU laws that affect his services. He is supported by the US government. Both the Trump administration and Musk accuse the Europeans of censorship and suppression of freedom of expression.
Elon Musk is industrializing sexual harassment
The Grok scandal is not a technical blip, but a symptom of what happens when generative AI is embedded in a social network. The ability to manipulate images directly on the platform shifts the line between expression and assault – and normalizes digital degradation as a function.
What is striking is not only the failure of protective mechanisms, but also the shift in responsibility. While the AI itself names legal categories such as “failure” or even criminal content, the company management relies on culture war rhetoric. Criticism is not addressed in terms of content, but rather politicized: media lies, laws suppress, regulation is considered censorship.
Grok is an example of Musk’s understanding of platform power. X is not a neutral space, but an instrument with which political interpretive sovereignty is tested and shifted. In this sense, the sexualized AI images are not an industrial accident, but rather collateral damage from a system that rewards provocation and only recognizes boundaries when courts draw them.
Voices
- Wolfram Weimer, Minister of State to the Federal Chancellor and Federal Government Commissioner for Culture and Mediatold Reuters: “What we are currently observing at
- xAI boss Elon Musk even actively fueled the wave of bikini manipulation. On his X account, he had Grok replace a meme of actor Ben Affleck with a picture of himself in a bikini. He does not want to take responsibility for other users misusing the function: “Anyone who uses Grok to create illegal content must expect the same consequences as if they uploaded illegal content.”
- In contrast to Musk, affected users do not enjoy the function at all. Samantha Smith, freelance journalist and commentatoris one of numerous victims. She told the BBC: “It wasn’t me being shown naked, but she looked like me and it felt like it was me. It was just as hurtful to me as if someone had actually posted a nude or bikini photo of me.”
Who controls AI?
What Elon Musk stages as humor and “radical freedom” meets a wall of paragraphs in Europe. Investigations in France, clear threats of punishment in Austria and the strict requirements of the AI Act and Digital Services Act put strict limits on the urge to play. Because: Deepfakes are not a meme, but a legal risk.
The pressure on X and xAI is likely to continue to increase, not only legally but also economically. Platforms that ignore systemic risks risk high penalties, restrictions or, in extreme cases, functional restrictions. Ironically, in the midst of the scandal, Grok announced new billion-dollar funding. However, this is likely to provide less of a tailwind than a race against regulation.
Ultimately, the question is who controls the power of generative AI: developers or society? Europe’s answer is increasingly law instead of gag and responsibility instead of excuse. Whether Musk accepts this course or continues to be confrontational could set a precedent.
Also interesting:
