Europe Strikes Grok: EU Targets AIs That Undress

Europe Reacts to the Grok Scandal

The recent scandal surrounding Grok has sent shockwaves across Europe, prompting Brussels to take drastic measures. The European Union is no longer content with merely sanctioning the dissemination of illegal content; it is now targeting the technologies that enable their creation. This new approach could transform the landscape of artificial intelligence.

Deepfakes, which have garnered significant attention this year, have also revealed their dark side. Grok allowed some users to generate intimate images of women without their consent, and even images of children were circulated. A simple prompt was enough to create these problematic contents, sparking an intense debate about the dangers of AI. In response to this situation, the EU has decided to directly target the tools responsible for these abuses.

Grok, Catalyst for European Response

The scandal gained traction on the platform X, where Grok enabled the creation of nude images of women and minors without any restrictions. The victims, including three teenage girls, have filed a lawsuit against xAI, the company behind Grok, for producing hyper-realistic pornographic images from their actual photos.

Although xAI eventually disabled this feature, public outrage has compelled European decision-makers to respond swiftly. According to Bloomberg, this scandal has accelerated an ongoing legislative process, shifting the focus from merely sanctioning abuses to eradicating their technological source.

Europe Expands Its Scope

On March 18, 2026, the LIBE committee in Brussels approved an amendment to the AI Omnibus, prohibiting systems capable of producing realistic sexual images without consent. This measure specifically targets identifiable individuals.

Until now, legislation primarily focused on content and its dissemination. Now, the emphasis is on the tools themselves, imposing technical constraints on developers and platforms to prevent the creation of such visuals. This preventive approach marks a turning point from traditional punitive measures.

The text, politically validated on March 11, 2026, by the Parliament and the Council, still needs to be voted on in a plenary session. Adjustments are possible, but sanctions could reach up to 7% of the global revenue of the companies involved.

Adjustments Under Pressure for the AI Act

This amendment is part of the AI Omnibus, aimed at adjusting the AI Act, the major European law on AI. The goal is to simplify certain rules, but discussions remain tense. Industry stakeholders fear that these technical constraints may stifle innovation, with a risk of deferring some obligations until 2027 or 2028 for high-risk applications.

Despite these concerns, Europe maintains its position. It aims to regulate generative AIs without hindering their development. Since 2024, a directive has already penalized violence against women, including that perpetrated by AI, and the Digital Services Act mandates the rapid removal of illegal content.

However, these measures show their limitations, as they intervene only after the content has been disseminated. Once online, images circulate rapidly, and their removal is insufficient to repair the damage caused.

The Grok Case, a Lesson for the Future

Although the amendment is still awaiting final validation, the Grok case in Europe has triggered investigations by the EU and the UK. Elon Musk's company, xAI, could face significant financial penalties, sending a strong signal to the entire sector.

Companies must now integrate safeguards from the design phase of their tools to avoid major legal risks. This case redefines priorities, placing responsibility at the heart of the development of intelligent automation solutions.

The tightening of regulations could also influence practices beyond European borders. The EU hopes, as with the GDPR, to impose its standards on a global scale. Developers will need to anticipate these requirements, integrate robust control systems, and limit certain uses from the outset.