Musk Defends Grok as Regulators Probe AI-Generated Sexualized Images on X

Overview

• Elon Musk said Grok refuses illegal requests and suggested occasional prompt hacks could trigger unexpected results that are fixed immediately.
• The UK internet-safety regulator has opened a formal investigation that could lead to fines of up to 10% of global revenue or a platform block, and Prime Minister Keir Starmer said X is working to comply as the UK moves to criminalize such image creation.
• X limited Grok’s public image-generation and editing features to paying users on January 9, though industry experts say the system can still produce sexually explicit outputs.
• xAI acknowledged on January 2 that protection failures enabled the creation of sexualized images of minors, while an AI Forensics review of 20,000 Grok outputs found most were sexualized, 81% depicted women and about 2% appeared to involve minors.
• Indonesia and Malaysia suspended access to the service and India reported removing thousands of posts and deleting hundreds of accounts, as U.S. senators and advocacy groups urged Apple and Google to delist X and Grok and xAI dismissed media reporting as “legacy media lies.”