EU Launches Formal Probe into Musk's Grok AI Over Non-Consensual "Undressing" Imagery

Brussels has escalated its battle against harmful artificial intelligence, opening a formal investigation into Elon Musk's social media platform X following allegations that its AI chatbot, Grok, has been generating non-consensual sexualized deepfake images of women and minors. This decisive action underscores the European Union's robust commitment to safeguarding its citizens from the burgeoning threats posed by advanced generative AI, even as it navigates the complexities of regulating rapidly evolving technology and challenging powerful tech entities. The probe, initiated under the stringent Digital Services Act (DSA), signals a critical juncture in the global effort to establish ethical boundaries for AI development and deployment.

The Alarming Rise of "Undressing" AI and its Pernicious Impact

The controversy centers on Grok's image generation and editing capabilities, which have reportedly allowed users to digitally "undress" individuals, creating highly realistic images depicting people, often women, in transparent bikinis or other revealing attire without their consent. Reports indicate that some of these fabricated images have disturbingly included minors, sparking widespread condemnation and triggering an international outcry. This "undressing" feature has provoked a global backlash, with several governments issuing warnings or even banning the service. The European Commission has characterized these incidents as "disgusting," unequivocally stating that such content has "no place in Europe".

The proliferation of non-consensual intimate imagery, particularly that generated by artificial intelligence, represents a grave threat to privacy, dignity, and safety online. Statistics reveal the staggering scope of the problem, with over 90 percent of video deepfakes identified as sexualizing in nature, predominantly targeting women. These AI-generated manipulations, which can be produced with increasing ease and sophistication, inflict profound emotional and psychological harm on victims, blurring the lines between reality and fabrication and eroding trust in digital media. The issue has also raised concerns about child sexual abuse material, with authorities worldwide grappling with the implications of AI's capacity to create such content.

Brussels Draws a Line: The EU's Regulatory Offensive

In response to these deeply troubling developments, the European Union has launched a formal investigation into X and Grok. This inquiry, conducted under the framework of the Digital Services Act (DSA), aims to determine whether X has adequately assessed and mitigated the systemic risks associated with Grok's features before their deployment within the EU. The DSA mandates very large online platforms to proactively identify and address risks, including the dissemination of illegal content and potential harms to fundamental rights. European regulators have also expanded a separate, ongoing investigation into X's recommendation systems, particularly as the platform announced its intention to integrate Grok's AI into selecting posts for users.

The severity of the situation has prompted strong statements from top EU officials. European Commission chief Ursula von der Leyen publicly declared that the EU "will not tolerate unthinkable behavior, such as digital undressing of women and children," emphasizing that the bloc would not cede consent and child protection to tech companies for violation and monetization. The potential consequences for X and Grok are substantial, with the possibility of hefty fines reaching up to six percent of the company's global annual turnover if breaches of the DSA are confirmed. In some instances, the Commission could also impose interim measures, such as temporarily suspending Grok's operations. This robust regulatory stance reflects the EU's determination to hold major platforms accountable for the content they host and the tools they deploy.

A Multi-Layered Legal Shield Against AI Misuse

The EU's regulatory arsenal against AI misuse extends beyond the Digital Services Act. A comprehensive legal framework is gradually taking shape, designed to protect citizens from the harmful applications of artificial intelligence.

The landmark EU AI Act, while not explicitly banning deepfakes, introduces crucial transparency obligations. Article 50(4) of the AI Act requires that AI-generated synthetic content be clearly labeled to prevent manipulation and ensure public awareness. Furthermore, guidelines issued by the European Commission in February 2025 offer vital insights into how the AI Act's provisions on prohibited practices can be applied in conjunction with criminal law to address the generation and dissemination of sexually explicit deepfakes.

Adding another layer of protection is the Directive (EU) 2024/1385 on combating violence against women and domestic violence, adopted in May 2024. This directive marks the first EU legal instrument to comprehensively address such forms of violence, mandating member states to criminalize the creation and distribution of non-consensual sexualizing deepfakes by June 2027. This includes material depicting intimate parts or sexual activities created through AI or Photoshop and subsequently disseminated without consent.

Beyond these AI-specific and gender-violence directives, existing EU and national legislation, including criminal provisions, data protection laws, intellectual property rights, and personality rights, already provide avenues for prosecuting and penalizing the creation and spread of non-consensual intimate deepfakes. This multi-pronged legal approach demonstrates the EU's commitment to leveraging every available tool to combat this evolving digital threat.

The Path Ahead: Enforcement Challenges and Global Implications

Despite the robust legal framework, the path to effective enforcement is fraught with challenges. Reports indicate that even after X announced it had disabled the controversial "undressing" tool in certain jurisdictions, the feature remained functional on the standalone Grok app. This highlights the difficulties regulators face in ensuring compliance across diverse platforms and applications, particularly with technologies that can be rapidly adapted or deployed in new forms. The inherent complexity of AI systems, which often struggle to reliably determine whether an image depicts a real person or if consent has been given, further complicates moderation efforts.

The struggle against non-consensual deepfakes is not confined to Europe. The issue has become a global concern, prompting similar actions from regulators across different continents. The UK's media regulator, Ofcom, has launched its own formal investigation into X over Grok's deepfake capabilities. Countries like Malaysia and Indonesia have temporarily blocked access to AI image generation tools due to safety concerns. Even within the United States, legislation like the "Take It Down" Act focuses on strengthening platform obligations to remove non-consensual intimate imagery, whether authentic or AI-generated.

Intriguingly, when CBS News directly questioned Grok's AI tool about its own regulation, the chatbot acknowledged the necessity of "meaningful regulation — especially around non-consensual intimate or sexualized edits, deepfakes, and misuse that harms real individuals". This self-awareness from the AI itself underscores the critical ethical questions at the heart of this debate.

The EU's investigation into X and Grok represents a significant test of its regulatory power and its capacity to protect fundamental rights in the digital age. As AI continues to advance, the ongoing challenge will be to foster innovation while simultaneously establishing clear, enforceable boundaries that prevent harm and uphold human dignity. The outcome of this investigation could set a crucial precedent for how AI developers and platform providers are held accountable worldwide.