Elon Musk’s artificial intelligence chatbot xAI continues to generate sexualized images of people even when explicitly warned that the subjects do not consent, according to a Reuters investigation. The findings raise fresh concerns about the effectiveness of safeguards announced by Musk’s social media platform X following global backlash over nonconsensual AI-generated imagery.
Reuters reporters conducted controlled tests of Grok after X introduced new limits on its public image-generation features. While the public-facing account reduced the volume of explicit output, the chatbot itself continued to produce sexualized images when prompted, including in cases where reporters warned that the subjects were vulnerable, would be humiliated, or had explicitly not consented. In dozens of test cases, Grok generated such images despite these warnings.
Rival AI systems from OpenAI, Alphabet and Meta declined to generate similar content and instead issued ethical or safety warnings. Regulators in Europe, Britain and the United States are examining whether Grok’s behavior could violate online safety, privacy, and consumer protection laws, particularly where minors or nonconsensual imagery are involved.
X and xAI did not provide detailed responses to Reuters’ findings, offering only brief statements dismissing media criticism. Legal experts warn that continued failures to prevent such content could expose the companies to fines, enforcement actions, or criminal liability in some jurisdictions.
