The European Commission has initiated a formal investigation into Elon Musk’s X platform, focusing on the social media company’s failure to prevent its Grok AI chatbot from generating sexually explicit deepfake images. The inquiry, launched on January 26, 2026, arises amid growing concerns about digital safety and compliance with EU regulations.
Regulators will scrutinize whether X conducted adequate risk assessments prior to deploying Grok’s image-generation capabilities across European markets. This investigation falls under the Digital Services Act (DSA), which requires large online platforms to implement robust safeguards against the distribution of illegal content.
Research conducted by the Center for Countering Digital Hate revealed that Grok produced approximately 3 million sexualized images within just two weeks of launching its image-editing feature. Alarmingly, about 23,000 of these images appeared to depict minors, according to the watchdog group’s findings.
In response to widespread international criticism, X restricted Grok’s image-editing functionalities to paying subscribers in late January and implemented geographic blocks in areas where such content contravenes local laws. However, officials from the European Commission asserted that these measures did not adequately address systemic risks associated with the platform.
“Sexual deepfakes of women and children represent a violent, unacceptable form of degradation,” stated Henna Virkkunen, the EU Commission’s executive vice-president for tech sovereignty, security, and democracy. “This investigation will determine whether X fulfilled its legal obligations under the DSA or treated European citizens’ rights as collateral damage.”
This latest inquiry builds on an ongoing DSA case against X that commenced in December 2023. That earlier investigation resulted in a €120 million penalty imposed on the platform in December 2025 for misleading verification practices and restrictions on research access.
Potential penalties for violations of the DSA could reach 6% of a company’s global annual revenue, with the Commission also holding the power to impose interim operational restrictions during the investigative process. Such actions reflect the EU’s commitment to enforcing stricter standards on digital platforms, particularly in light of the troubling implications of AI-generated content.
Earlier this year, multiple Asian countries temporarily blocked access to Grok due to concerns over child safety. Nations such as Indonesia, Malaysia, and the Philippines reinstated access after xAI, the parent company of X, instituted additional content safeguards.
In a parallel development, the UK’s media regulator, Ofcom, launched its own investigation into X in January 2026, focusing on the potential generation of child sexual abuse material. EU officials will also analyze X’s content recommendation systems, particularly following the company’s announcements about enhancing Grok’s integration for content filtering. Regulators have voiced concerns that AI-driven recommendations could facilitate the rapid dissemination of illegal content.
“This content has no place in Europe,” stated EU Commission spokesperson Thomas Regnier during a press conference in Brussels addressing X’s controversial “Spicy Mode” feature. “We will not tolerate unthinkable behavior such as the digital undressing of women and children.”
This investigation involves close coordination with Ireland’s national Digital Services Coordinator, Coimisiún na Meán, and will allow for evidence collection through document requests, interviews, and potential platform inspections. Despite increasing regulatory pressure, xAI managed to secure $20 billion in Series E funding on January 6, 2026, attracting investments from notable entities such as the Qatar Investment Authority, MGX, Nvidia, and Cisco.
This funding round comes as X faces mounting scrutiny regarding Grok’s capabilities and its implications for digital safety. The outcome of this investigation represents a significant test of the EU’s enforcement of the Digital Services Act against the backdrop of rapidly evolving AI technologies.
See also
UC Berkeley Team Reveals Generative Meta-Model Achieving LLM Interpretability with 1B Activations
OpenAI Discontinues GPT-5, GPT-4o Models to Focus on User-Favored Upgrades
Google Unveils Gemini Embedding Models, Enhancing AI Search and Recommendations
Caltech Researchers Achieve 11x Improvement in Carbon Capture Simulations with Fun-DDPS
