Two prompts used to generate the material on Grok were rejected by ChatGPT, Meta AI and OpenAI’s Anthropic’s Claude as inappropriate during WIRED testing. Google Gemini did produce a photo of one of the stars held in the giant’s hand, although it rejected another suggestion. Google declined to comment.
One of the Grok Imagine videos, which was also posted on X, appeared to feature Ashley St. Clair dressed up to dance in a bikini. St. Clair was previously in a relationship with Musk and is the mother of one of his children. in January she started legal action against xAI after its alleged sexual imitations appeared on X. After WIRED contacted X, the post was removed from the social media platform for violating its policies.
Legal representatives of St. Clair on X did not immediately respond to a request for comment.
“Elon Musk deliberately added a perverse feature to his platform that helps users undress women and children with a single click, regardless of the foreseeable harm it may cause,” says Imran Ahmed, CEO and founder of the Center for Countering Digital Hate, which in January estimated with a high degree of certainty, which Grok created 3 million sexually explicit images allegedly depicting over 20,000 children. “Now it appears that vulgar content is still stored on Grok, including photos ridiculing the mother of Musk’s child,” he added.
Unlike other generative AI systems from OpenAI and Google, Musk’s Grok and xAI did not back away from generally allowing sexual content, having previously introduced “Spicy” and “Unhinged” modes, and initially included fewer guardrails. Musk has he stated Grok “presumably [to] “allow upper body nudity of imaginary adult humans” and be consistent with what viewers may see in R-rated films. Latest terms of use of xAI services they claim that the system can respond with “sexual situations”. However, the company’s documentation says it does not allow people to use its systems to “cause harm or engage in abuse.”
Other Grok Imagine videos viewed by WIRED feature women who are believed to be entirely AI-generated, undressing or engaging in sexual acts – some completely explicit. User prompts for many of the videos do not necessarily directly describe sex acts, but describe them in a roundabout way, likely an attempt to circumvent security measures implemented on the Grok platform.
Multiple researchers tell WIRED that since January, X and Grok’s changes appear to have made it more difficult to “nudify” or “undress” images of real people. The number of these photos posted on X appears to be decreasing in recent months. On Reddit and one dedicated forum dedicated to false claims about artificial intelligence, users have complained about increased moderation by SpaceX-owned companies.
Nevertheless, in May, SpaceX warned potential investors that it had set aside $530 million to resolve ongoing legal complaints, including those related to Grok. “Because these modes may be more dismissive and stringent than our standard offerings, they involve greater risks, including loss of reputation, the generation of potentially explicit content and misinformation or fraudulent results, the potential for non-consensual or exploitative use of images, intellectual property infringement, or content that could be perceived as exploitative, harmful, harassing, abusive, or discriminatory,” the May filing reads.