France24 reports that the European Union has moved closer to banning one of the most harmful uses of AI.

Lawmakers have approved new rules targeting AI systems that create sexualized deepfakes without consent after Grok’s stunts in January. 

At the center of the debate is xAI’s chatbot, Grok. The tool faced criticism after users generated

The European Parliament voted to prohibit so-called “nudifier” applications. These tools use AI to create or manipulate images into explicit content. 

Specifically, the ban targets AI that generates fake nude or sexual images, content that resembles identifiable individuals, and images created without clear consent. 

In other words, any AI tool that produces explicit images of real people without permission will face restrictions.

Lawmakers, however, stated that AI tools with effective safeguards will not be affected. If a platform actively prevents misuse, it may remain compliant.

The Origin

Earlier this year, users on X shared AI-generated explicit images. Many victims had no knowledge or control over the content.

For example, one widely reported case involved a young woman whose images were manipulated without consent. 

Although the images were fake, they appeared highly realistic. Consequently, the emotional impact was severe.

Such cases increased public pressure. X announced updates to limit harmful content, especially involving minors and women. 

Nevertheless, critics argued that stronger safeguards should have come sooner.

Also read: Grok’s Non-consensual Images On X Draw’s Global Scrutiny

Legislative Process 

The proposal has received strong support from EU lawmakers. In fact, most member states have already approved the direction of the rules.

Next, policymakers will finalize negotiations between institutions, agree on precise legal language, and formally adopt the regulation. 

Officials expect the process to move quickly as a general agreement already exists. Therefore, final approval appears likely. 

AI Regulations

Alongside the deepfake ban, lawmakers addressed other AI rules, supporting measures to delay implementation of high-risk AI systems.

These systems include technologies that may affect public safety, health outcomes, and fundamental rights 

Originally, implementation was set for 2026, but timelines may change.