Grok AI generated 3 million sexualized images in 11 days: Study
A new study found that Grok, an xAI chatbot integrated into X (release year not specified in the source), created about three million sexualized images in just 11 days—including around 23,000 of children—after a one-click image-editing feature on X was rolled out recently.
Researchers analyzed a sample of posts to uncover just how fast and widespread this became.
Users created images at lightning speed
People used simple prompts like "put her in a bikini" or "remove her clothes" on real photos, often without consent.
The content was generated at a wild rate—about 190 images every minute.
Researchers used smart detection tools and manual checks to confirm the numbers.
Big backlash sparks urgent changes
The scale of the problem drew criticism and pressure from UK authorities and prompted legal action in several countries.
X responded by restricting image edits to paying users and later blocked undressing features for real people's photos.
Experts warn that nonconsensual sexual imagery can seriously harm victims, especially women and girls.
Why it matters: Time for stronger safeguards
The study says we need better safety features built into these AI tools from the start—not just fixes after things go wrong.
It's a wake-up call for stricter rules and smarter design so tech like this can't be misused so easily again.