{"id":56390,"date":"2026-06-12T15:53:11","date_gmt":"2026-06-12T19:53:11","guid":{"rendered":"https:\/\/overcentral.com\/en\/?p=56390"},"modified":"2026-06-12T15:53:11","modified_gmt":"2026-06-12T19:53:11","slug":"grok-hosts-sexualized-deepfakes-safety-violations","status":"publish","type":"post","link":"https:\/\/overcentral.com\/en\/grok-hosts-sexualized-deepfakes-safety-violations\/","title":{"rendered":"Grok Hosts Sexualized Deepfakes of Famous Women Despite Safety Pledges"},"content":{"rendered":"

Grok, the generative AI<\/a> system developed by Elon Musk’s xAI, continues to host and distribute sexualized deepfake content featuring famous women, despite public safety commitments from the company. Two prompts that successfully generated explicit material on Grok were rejected by OpenAI’s ChatGPT, Meta AI<\/a>, and Anthropic’s Claude as inappropriate when tested. Google’s Gemini produced one image of a celebrity but declined a second prompt. The findings underscore a persistent gap between safety pledges and actual enforcement on Musk’s platform.<\/p>\n

Grok Deepfake Incident Involving Ashley St. Clair Highlights Platform Risks<\/h2>\n

One video created using Grok Imagine and subsequently posted to X depicted Ashley St. Clair, who was previously in a relationship with Elon Musk<\/a> and is the mother of his child, altered to appear dancing in a bikini. In January, St. Clair initiated legal action against xAI after sexualized deepfakes of her allegedly appeared on X. The post was removed from the platform only after WIRED<\/a> contacted X, with the company citing a violation of its rules.<\/p>\n

Carrie Goldberg, St. Clair’s attorney and a prominent advocate against nonconsensual sexualized imagery, described the creation of such content featuring celebrities as a “historic” development. “So there was instant publication of the digital nude images,” Goldberg stated. “And not only that, but the world’s richest man was promoting the nudification technology to his 240 million followers on X, monetizing it, and doing so on a platform that is for ages 13+.”<\/p>\n

CCDH Estimates Millions of Sexualized Images Created Through Grok<\/h2>\n

Imran Ahmed, CEO and founder of the Center for Countering Digital Hate<\/a>, echoed Goldberg’s concerns. “Elon Musk knowingly added a perverse feature to his platform that helps users undress women and children at the click of a button, with no regard for the predictable damage it would cause,” Ahmed claimed. In January, CCDH estimated with high confidence that Grok created 3 million sexualized images, allegedly including more than 20,000 depicting children. “Now it appears explicit content is still being hosted on Grok and shared on X, including images ridiculing the mother of Musk’s child,” Ahmed added.<\/p>\n

Grok’s Guardrails Remain Weak Compared to Competing AI Systems<\/h2>\n

Unlike generative AI systems from OpenAI and Google, Grok and xAI have not stepped back from permitting sexual content. The platform introduced “Spicy” and “Unhinged” modes with fewer safety guardrails. Musk has stated Grok is “supposed to allow upper body nudity of imaginary adult humans” and be consistent with content found in R-rated movies. The most recent xAI terms of service indicate the system may respond with “sexual situations,” though the company’s documentation states it does not allow use of its systems for “causing harm or engaging in abusive activity.”<\/p>\n

Other Grok Imagine videos reviewed show women, likely entirely AI-generated, undressing or involved in explicit sexual acts. User prompts for many videos describe sexual scenarios indirectly, a tactic that appears designed to circumvent safeguards deployed on the Grok platform.<\/p>\n

How Are AI Deepfake Platforms Like Grok Being Used to Create Nonconsensual Sexualized Images?<\/h2>\n

Generative AI systems can be prompted to create realistic images of individuals without their consent. Grok’s relatively permissive content policies and limited safety guardrails have enabled users to generate sexualized deepfakes of real people by describing scenarios in indirect language. The images can then be instantly published on X, reaching millions of viewers. Unlike platforms that proactively block such prompts, Grok’s approach has allowed the creation and distribution of this material at scale.<\/p>\n

Recent Enforcement Changes Show Limited Progress<\/h2>\n

Multiple researchers report that since January, changes introduced by X and Grok have made it harder to create “nudification” or “undress” images of real people. The number of such images posted to X has decreased in recent months. On Reddit and dedicated AI deepfake forums, users have complained about increased moderation from xAI. However, the continued presence of explicit content and the platform’s foundational permissiveness indicate that the problem remains unresolved.<\/p>\n

What Affected Users Should Do Now<\/h2>\n

Individuals who discover nonconsensual sexualized images of themselves on Grok or X should immediately document the content by taking screenshots and recording the URLs, then report the material through X’s reporting tools and to relevant legal authorities. Affected users should consider consulting with an attorney who specializes in digital privacy and nonconsensual imagery cases. For broader protection against deepfake misuse, users should monitor their digital footprint regularly, restrict public sharing of high-quality personal images, and enable two-factor authentication on all social media accounts. When choosing an AI platform, consumers should prioritize services that implement robust content moderation and enforce clear policies against nonconsensual sexualized imagery. Users in the US, UK, Australia, and Canada should also familiarize themselves with local laws regarding deepfake content, as legal frameworks in these jurisdictions increasingly provide avenues for recourse against such abuse.<\/p>\n","protected":false},"excerpt":{"rendered":"

Grok, the generative AI system developed by Elon Musk’s xAI, continues to host and distribute sexualized deepfake content featuring famous women, despite public safety commitments from the company. Two prompts that successfully generated explicit material on Grok were rejected by OpenAI’s ChatGPT, Meta AI, and Anthropic’s Claude as inappropriate when tested. Google’s Gemini produced one […]<\/p>\n","protected":false},"author":7,"featured_media":73770,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/iili.io\/CCOSgWu.jpg","fifu_image_alt":"Grok Hosts Sexualized Deepfakes of Famous Women Despite Safety Pledges","footnotes":""},"categories":[349],"tags":[],"class_list":["post-56390","post","type-post","status-publish","format-standard","has-post-thumbnail","category-articles"],"fifu_image_url":"https:\/\/iili.io\/CCOSgWu.jpg","fifu_image_alt":"Grok Hosts Sexualized Deepfakes of Famous Women Despite Safety Pledges","_links":{"self":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/posts\/56390","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/users\/7"}],"replies":[{"embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/comments?post=56390"}],"version-history":[{"count":0,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/posts\/56390\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/media\/73770"}],"wp:attachment":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/media?parent=56390"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/categories?post=56390"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/tags?post=56390"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}