Overview
- New analysis from the Center for Countering Digital Hate estimates Grok produced about 3 million sexualized images in 11 days, including roughly 23,000 appearing to depict children, based on extrapolating a 20,000‑image sample from 4.6 million outputs.
- Malaysia lifted its Grok suspension after receiving confirmation of additional safeguards, while the Philippines said it will end its short ban once xAI removes local deepfake and pornographic functions; Indonesia’s temporary block remains in place.
- xAI and X limited image editing to paying users, added technical barriers, and announced geoblocking in some jurisdictions, but reporting finds the standalone Grok app and prompt workarounds can still create problematic images, with some posts remaining accessible by direct URL.
- Regulatory and legal pressure is intensifying, including probes by California and Arizona attorneys general, UK Ofcom action under the Online Safety Act, EU scrutiny under the Digital Services Act, and French and Australian inquiries, alongside petitions urging Apple and Google to pull the app.
- India’s electronics ministry is weighing whether to treat Grok as an X account for intermediary liability and considering a Grok‑only ban, while lawsuits such as Ashley St. Clair’s case and evidence of celebrity targets underscore potential harms and accountability risks.