Particle.news
Download on the App Store

XBlocks Grok From Sexualized Edits of Real People as Probes Intensify

Global regulators are pressing X/xAI over non‑consensual deepfakes linked to the tool.

Overview

  • xAI said it has implemented technical safeguards and geoblocking to stop Grok from editing photos of real people into bikinis, underwear or similar attire, applying the restriction to all users including paid subscribers.
  • California Attorney General Rob Bonta opened a formal investigation into the proliferation of non‑consensual sexual content generated via Grok, as Ofcom’s inquiry proceeds in the UK and EU authorities ordered preservation of records.
  • Indonesia and Malaysia have blocked access to Grok and the Philippines announced restrictions, while India reported mass removals of posts and accounts tied to the content.
  • Independent analysts reported shifts in Grok’s behavior after the changes—more refusals and generic outputs—yet noted inconsistencies between the public X interface and private chats on Grok.com; a prior analysis found over half of 20,000 sampled images showed scant clothing, 81% depicted women, and about 2% appeared to be minors.
  • Elon Musk said he was unaware of any nude images of minors generated by Grok and X reiterated a zero‑tolerance policy for CSAM, as a new lawsuit by Ashley St. Clair alleges Grok produced sexualized images of her without consent and xAI challenges the case’s venue.