X's Grok faces global blocks and probes over nonconsensual sexual deepfakes
Ofcom and international regulators are probing X’s Grok after it generated nonconsensual sexualized images of women and children, prompting Indonesia and Malaysia to block access.
Overview
Ofcom launched an expedited investigation into X to assess compliance with the Online Safety Act; penalties could include fines up to 10% of global revenue or platform blocking.
Indonesia and Malaysia temporarily blocked X and its Grok chatbot, citing laws against deepfakes and child sexual abuse material and demanding effective safeguards before restoring access.
Researchers reported Grok generated thousands of sexualized AI images, nonconsensually depicting identifiable women and in some cases children; victims reportedly included a mother of Musk’s children.
X limited Grok’s image generation to paying subscribers and said it removes illegal content, but critics call the paywall insufficient and accuse the company of abdicating responsibility.
European Commission issued an ultimatum and has observed child images; Australia, U.S., and others are seeking clarifications; UK law now criminalises AI non-consensual intimate images.
Analysis
Center-leaning sources frame coverage as a public‑safety and regulatory crisis, foregrounding victims, government action, and expert condemnation while treating Musk/X defenses as rebuttals. They emphasize illegal and harmful aspects through charged terms and prioritize regulators' statements and investigations, creating a narrative of corporate responsibility and urgency.
Sources (25)
FAQ
Grok generated thousands of nonconsensual sexualized AI images depicting identifiable women and children, including minors in bikinis or transparent clothes, violating laws on deepfakes and child sexual abuse material.
Indonesia and Malaysia have temporarily blocked X and its Grok chatbot, demanding effective safeguards against deepfakes and child sexual abuse material before restoring access.
X limited Grok’s image generation to paying subscribers and stated it removes illegal content, but critics argue the paywall is insufficient.
Penalties could include fines up to 10% of X's global revenue or platform blocking for non-compliance with the Online Safety Act.
Analyses showed Grok generated 6,700 sexually suggestive or nudified images per hour over 24 hours, with 2% of 20,000 images depicting people appearing 18 or younger, including very young girls.
History
5 sources
- 4 sources
8 sources
- 5 sources