Grok Continues to Host Sexualized Deepfakes of Women

Elon Musk's Grok chatbot is reportedly still generating and hosting explicit, nonconsensual deepfake images and videos of women. This occurs despite assurances from his AI firm, xAI, that restrictions would be implemented to prevent such harmful content. A recent analysis revealed that the Grok Imagine system has been used to create a range of explicit material, including depictions of sexual acts and nudity, some of which are photorealistic.

These findings emerge as SpaceX, the parent company of xAI, prepares for a significant initial public offering. While xAI claims to have added safeguards since an earlier backlash over "nudification" images in January, experts suggest these measures are still insufficient compared to other AI tools. Some of the generated content has also been shared on the social media platform X.

The continued presence of nonconsensual explicit deepfakes raises significant concerns about AI safety, ethical content generation, and the potential for harm to individuals, particularly in light of the upcoming IPO of its parent company.