Nonconsensual explicit images are still being created and shared using Grok, despite new restrictions. Legal and regulatory pressure is mounting as SpaceX prepares for its IPO. xAI’s safeguards are under question.
Months after xAI pledged to curb the creation of nonconsensual sexualized deepfakes, Grok’s generative AI system continues to produce and host explicit images and videos of women without their consent. This ongoing issue comes as SpaceX, the parent company of xAI, readies for one of the largest IPOs in history.
WIRED’s analysis of public Grok Imagine links revealed dozens of sexualized AI-generated images and videos, including depictions of celebrities and at least one politician. Some content was photorealistic, while others used animated styles. Many of these creations were later shared on X, though most generations are not public by default.
Compared to other generative AI platforms, Grok has fewer safety guardrails. Henry Ajder, a deepfake expert, noted that while some changes have been made since the backlash over “nudification” earlier this year, Grok still lags behind mainstream tools in preventing abuse.
xAI has faced lawsuits and regulatory scrutiny since January, when Grok was used to generate a surge of sexualized images, including some allegedly involving minors. The company claims to have implemented new safeguards and maintains that child sexual abuse material is banned. However, neither xAI nor X responded to WIRED’s requests for comment regarding the explicit content still hosted on Grok.com.
After WIRED contacted the companies, many explicit images and videos were removed from Grok.com and X for policy violations. In April, X’s safety account reiterated its prohibition on generating nonconsensual explicit deepfakes and using its tools to undress real people.
WIRED identified images and videos on Grok.com depicting multiple celebrities and U.S. Representative Alexandria Ocasio-Cortez. In some, women appeared in minimal clothing, held by a giant man’s hand, with prompts describing nonconsensual scenarios. When WIRED tested similar prompts on OpenAI’s ChatGPT, Meta AI, and Anthropic’s Claude, all rejected them as inappropriate. Google’s Gemini allowed one such image but rejected another; Google declined to comment.
One Grok Imagine video, also posted to X, showed Ashley St. Clair, a former partner of Elon Musk, altered to appear dancing in a bikini. St. Clair began legal action against xAI in January after sexualized deepfakes of her surfaced on X. The post was removed after WIRED’s inquiry. Her legal team did not respond to requests for comment.
Imran Ahmed, CEO of the Center for Countering Digital Hate, accused Elon Musk of knowingly enabling features that facilitate the undressing of women and children, estimating Grok created 3 million sexualized images, including over 20,000 of children. He claims explicit content remains on Grok and X, including images targeting the mother of Musk’s child.
Unlike OpenAI and Google, xAI has not fully restricted sexual content on Grok, previously introducing “Spicy” and “Unhinged” modes with fewer safety measures. Musk has stated Grok is intended to allow upper body nudity of imaginary adult humans, consistent with R-rated movies. xAI’s terms of service permit responses involving “sexual situations,” but prohibit harmful or abusive activity.
WIRED observed Grok Imagine videos showing AI-generated women undressing or engaging in sexual acts, sometimes described in indirect prompts to bypass safeguards. Researchers say recent changes have made it harder to create “nudification” images of real people, and such posts on X have declined. Users on Reddit and deepfake forums have complained about increased moderation.
Despite these changes, SpaceX disclosed in May that it has set aside $530 million for ongoing legal complaints, including those related to Grok. The company’s IPO filing warned of risks from “irreverent” modes, including reputational harm, explicit content, misinformation, and exploitative imagery.
In January, the Privacy Commissioner of Canada began investigating xAI and Grok, alleging violations of federal privacy law due to insufficient safeguards. xAI claims to have added new protections and proactive checks for infringing content, but the Commissioner remains unconvinced, stating xAI has not demonstrated the effectiveness of these measures.
Founded by Elon Musk in 2023, xAI has quickly become a major player in the generative AI space. As of 2026, the company employs over 1,000 people and has raised more than $2 billion in funding. Grok, its flagship chatbot, is integrated into X and has attracted millions of users, but ongoing legal and regulatory challenges continue to shape its future.
