Elon Musk's Grok chatbot has been found hosting nonconsensual and explicit deepfake images of women, months after promises to improve safety measures. The revelations come as SpaceX prepares for a major IPO, highlighting the ongoing challenges in regulating AI-generated content.

The analysis by WIRED uncovered numerous sexualized deepfakes featuring celebrities and politicians, with Grok Imagine producing highly realistic photorealistic images that even evade detection by other AI systems like ChatGPT and Google Gemini. Musk's xAI has faced lawsuits and scrutiny over the issue, but its response has been patchy at best.

Henry Ajder, an expert on deepfakes, comments: 'While Grok and X may have made some amendments to their model following backlash, they still fall short of industry standards.' The lack of robust safeguards means that explicit content continues to be hosted without the consent of those depicted.

The case with Ashley St. Clair, who previously filed legal action against xAI, highlights the severe impact such technology can have on individuals. Imran Ahmed from CCDH criticizes Musk for knowingly promoting this feature, which he claims will cause predictable damage to victims and their families.

While other AI systems like OpenAI and Google have backed away from allowing sexual content in general, Grok remains an outlier, with its 'Spicy' and 'Unhinged' modes contributing to the ongoing issue of nonconsensual and explicit deepfakes.