Explainer: Misuse of Elon Musk’s AI Tool Grok to Create Sexually Violent and Explicit Content

Elon Musk’s AI tool Grok, designed for image and video generation, has been found to be used for creating sexually violent and explicit content featuring women, according to recent research by AI Forensics, a Paris-based non-profit organization. The investigation uncovered approximately 800 pornographic images and videos generated using the Grok Imagine app, some of which were photorealistic and professionally produced. The content ranged from erotic imagery and suggestive poses to full nudity and sexual acts, including disturbing depictions such as a woman tattooed with “do not resuscitate” holding a knife between her legs. Notably, Grok was also used to manipulate an image of Renee Nicole Good, a woman recently killed by a U.S. Immigration and Customs Enforcement (ICE) agent, to undress her and portray her with a bullet wound in her forehead. These altered images appeared on X, the social media platform owned by Musk’s company xAI, which has integrated Grok.

AI Forensics was able to access these images because users created “sharing links” that were archived by the Wayback Machine, an internet archive service. While it remains unclear if the explicit content was widely distributed on X itself, the research highlighted that the explicitness of content generated by Grok far exceeds previous trends of bikini images on the platform. The misuse of Grok has sparked strong condemnation from political and social figures. UK Prime Minister Keir Starmer criticized the proliferation of AI-generated sexually explicit images of women and children on X, calling the content “disgraceful” and “disgusting,” and demanding urgent action from the platform to remove such material. Women’s rights groups, including the Fawcett Society, have also called on the UK government to impose stricter regulations on AI tools to prevent harm and humiliation of women.

The AI Forensics report analyzed 50,000 mentions of “@Grok” on X and 20,000 images generated by the tool over a week-long period. It found that about a quarter of the mentions were requests for image creation, with many prompts involving terms like “her,” “put,” “remove,” “bikini,” and “clothing.” More than half of the generated images depicted people in minimal attire, predominantly women who appeared to be under 30 years old. Alarmingly, around 2% of images seemed to show individuals aged 18 or younger. The report also cited an incident where a teenage girl requested Grok to alter a personal photo, which was then exploited by male users to generate offensive and inappropriate modifications, including dressing her as a Nazi and placing her in a bikini.

The controversy has raised significant concerns about the lack of safeguards in AI image generation tools and the ethical implications of their misuse. Critics argue that without robust regulation and oversight, AI technologies like Grok can be weaponized to produce harmful content that violates privacy, dignity, and legal boundaries. Musk’s xAI responded by stating that users creating illegal content with Grok will face consequences equivalent to those for uploading illegal material. However, the incident underscores the urgent need for clearer policies and stronger enforcement mechanisms to prevent AI-facilitated abuse.

In summary, the misuse of Grok highlights the broader challenges of AI governance, especially as these tools become more accessible and capable of producing realistic synthetic media. The case also illustrates the intersection of technology, social media, and legal frameworks, emphasizing the importance of proactive measures to protect vulnerable individuals from digital exploitation. As AI continues to advance, balancing innovation with ethical responsibility remains a critical priority for developers, platforms, and regulators alike.