Brazilian authorities have ordered Elon Musk’s social media platform, X (formerly Twitter), to immediately stop its artificial intelligence chatbot, Grok, from generating sexually explicit deepfake images.
The directive was issued on Wednesday by Brazil’s National Data Protection Agency (ANPD), the National Consumer Rights Bureau (Senacon), and the country’s chief prosecutor, following concerns that Grok was being used to create sexualised images of women and minors without consent.
According to the agencies, X must urgently implement safeguards to prevent the production of sexualised or eroticised content involving children, adolescents and non-consenting adults. The platform has been given five days to comply or face legal action and possible fines.
Authorities said X had earlier claimed it removed thousands of posts and suspended hundreds of accounts after receiving a warning in January. However, further checks revealed that users were still able to generate sexualised deepfakes using Grok.
Brazilian regulators criticised the company for what they described as a lack of transparency in its response to the issue.
The move comes amid growing international pressure on xAI, the company behind Grok, after the chatbot’s “Spicy Mode” feature enabled users to create explicit deepfake images using simple prompts. Countries such as Indonesia have already moved to block Grok entirely, while Britain and France have vowed to keep up regulatory pressure.
A report by the Center for Countering Digital Hate (CCDH) estimated that Grok generated millions of sexualised images of women and children within days of the feature being enabled.
X had announced earlier that it introduced measures to stop Grok from digitally undressing images of real people in countries where such practices are illegal, but Brazilian authorities said the safeguards remain ineffective.
