Elon Musk’s Grok Generated 23K Sexualized Images of Children, Says Watchdog

Elon Musk’s AI chatbot Grok produced an estimated 23,338 sexualized photos depicting kids over an 11-day interval, in accordance with a report launched Thursday by the Middle for Countering Digital Hate.

The determine, CCDH argues, represents one sexualized picture of a kid each 41 seconds between December 29 and January 9, when Grok’s image-editing options allowed customers to control photographs of actual folks so as to add revealing clothes and sexually suggestive poses.

The CCDH additionally reported that Grok generated practically 10,000 cartoons that includes sexualized kids, primarily based on its reviewed knowledge.

The evaluation estimated that Grok generated roughly 3 million sexualized photos complete throughout that interval. The analysis, primarily based on a random pattern of 20,000 photos from 4.6 million produced by Grok, discovered that 65% of the pictures contained sexualized content material depicting males, ladies, or kids.

“What we discovered was clear and disturbing: In that interval Grok grew to become an industrial-scale machine for the manufacturing of sexual abuse materials,” Imran Ahmed, CCDH’s chief govt informed The Guardian.

Grok’s transient pivot into AI-generated sexual photos of youngsters has triggered a world regulatory backlash. The Philippines grew to become the third nation to ban Grok on January 15, following Indonesia and Malaysia within the days prior. All three Southeast Asian nations cited failures to stop the creation and unfold of non-consensual sexual content material involving minors.

In the UK, media regulator Ofcom launched a proper investigation on January 12 into whether or not X violated the On-line Security Act. The European Commission mentioned it was “very significantly trying into” the matter, deeming these photos as unlawful underneath the Digital Companies Act. The Paris prosecutor’s workplace expanded an ongoing investigation into X to incorporate accusations of producing and disseminating little one pornography, and Australia started its own investigation too.

Elon Musk’s xAI, which owns each Grok and X—previously Twitter, the place lots of the sexualized photos had been mechanically posted)—initially responded to media inquiries with a three-word assertion: “Legacy Media Lies.”

Because the backlash grew, the corporate later implemented restrictions, first limiting picture technology to paid subscribers on January 9, then including technical boundaries to stop customers from digitally undressing folks on January 14. xAI announced it might geoblock the characteristic in jurisdictions the place such actions are unlawful.

Musk posted on X that he was “not conscious of any bare underage photos generated by Grok. Actually zero,” including that the system is designed to refuse unlawful requests and adjust to legal guidelines in each jurisdiction. Nevertheless, researchers discovered the first concern wasn’t absolutely nude photos, however relatively Grok putting minors in revealing clothes like bikinis and underwear, in addition to sexually provocative positions.

As of January 15, a few third of the sexualized photos of youngsters recognized within the CCDH pattern remained accessible on X, regardless of the platform’s acknowledged zero-tolerance coverage for little one sexual abuse materials.