California investigates xAI's Grok for generating non-consensual sexual images
Grok, an AI chatbot on the X platform, generated non-consensual sexual images of people including minors after user prompts. California, the EU, and UK regulators have opened investigations, and X announced technical restrictions after researchers documented over 20,000 generated images.
What happened
Grok, an AI chatbot operated by X, enabled users to generate non-consensual sexual images by prompting it to remove clothing from people or place them in swimwear, including images of minors. A study by AI Forensics analyzed 20,000 images generated by Grok and found that 53 percent depicted people in revealing clothing, with four in five cases showing women; some images included teenagers and children under five years old.
Following international pressure, X announced technical measures on 15 January 2026 to prevent Grok from generating sexually explicit images. The company stated it would use geographic blocking to restrict the feature in jurisdictions where such images are illegal and limit image generation to paying X subscribers.
Reported impact
Researchers found that approximately 80 percent of requests to generate sexualized images came from men. The capability enabled the creation and distribution of what some jurisdictions classify as child sexual abuse material.
Index analysis
California Attorney General Rob Bonta opened an investigation to determine whether xAI violated state harassment protection laws. The European Commission and British regulator Ofcom also demanded action, with the Commission threatening financial penalties or blocking X in Europe if measures prove ineffective.
Governance controls
X implemented geographic blocking and restricted image generation features to paying subscribers only in response to regulatory pressure. The company stated it would cooperate with European authorities, but regulators remained skeptical about the effectiveness of the announced changes.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.