The two signals used to generate content on Grok were rejected as unsuitable by OpenAI’s ChatGPT, Meta AI and Anthropic’s Cloud when tested by WIRED. Google’s Gemini produced an image of a celebrity holding the hand of a giant, although it rejected the second signal. Google declined to comment.
A Groc Imagine video, also posted on X, featured Ashley St. Clair dancing in a bikini. St. Clair was previously in a relationship with Musk and is the mother of his child. In January, she launched legal action against XAI after her sexual deepfakes allegedly appeared on X. After WIRED contacted Ax, the post was removed from the social media platform for violating its rules.
Legal representatives for St. Clair in the X case did not immediately respond to a request for comment.
Imran Ahmed, CEO and founder of the Center for Countering Digital Hate, claims, “Elon Musk knowingly added a perverted feature to his platform that helps users undress women and children at the click of a button, without regard to the potential harm it may cause.” “It now appears that explicit content is still being hosted on Grok and shared on X, including images ridiculing the mother of Musk’s child.”
Unlike OpenAI and Google’s other generative AI systems, Musk’s Grok and Musk has said that Grok is “considered.” [to] Allow upper body nudity of fictional adult humans” and be consistent with what viewers might see in R-rated movies. XAI’s latest terms of service state that the system can interact with “sexual situations.” However, the company’s documents say it does not allow people to use its system to “engage in harm or abusive activity.”
Other Grok Imagine videos seen by WIRED show women, presumably generated entirely by AI, taking off clothes or engaging in sexual acts — some completely explicit. User prompts for many videos do not necessarily directly describe sexual acts, but they describe them in a roundabout way – a possible attempt to circumvent the security measures deployed on the Grok platform.
Multiple researchers tell WIRED that changes introduced by Ax and Grok since January appear to have made it harder to create “nudity” or “undressed” images of real people. The number of these images posted on X has seen a decline in recent months. On Reddit and a dedicated AI deepfake forum, users have complained about increased moderation from the SpaceX-owned companies.
Still, in May, SpaceX warned potential investors that it had set aside $530 million to handle ongoing legal complaints, including those involving Grok. “Because these methods may be more abusive and harsh than our standard offerings, they present increased risks, including reputational damage, potentially explicit content and the generation of misinformation or misleading outputs, potentially non-consensual or exploitative imagery, intellectual property infringement, or content that could be viewed as exploitative, harmful, harassing, abusive or discriminatory,” its filing in May said.
<a href