Brazil orders X to 'immediately' block Grok sexualized deepfakes

Brazil on Wednesday told Elon Musk's social network X to stop its chatbot from creating sexually explicit images, the latest country to pressure the billionaire to fix the artificial intelligence tool.

Indonesia became the first country to block Grok entirely last month, while Britain and France said they would maintain pressure after the chatbot cranked out a flood of lewd photos of women and children.

X must "immediately implement appropriate measures to prevent the production, using Grok, of sexualized or eroticized content of children and adolescents, as well as adults who have not given their consent," said Brazil's chief prosecutor, the National Data Protection Agency (ANPD) and the National Consumer Rights Bureau (Senacon.)

The agencies gave the platform five days to comply with the order or risk legal action and fines.

X claimed it had deleted thousands of posts and suspended hundreds of accounts after Brazil issued a warning last month, Brazilian authorities said.

But checks found that Grok users were still able to generate sexualized deepfakes, they added.

They slammed X for "not being transparent in its response."

X announced measures on January 15 to prevent Grok from undressing images of real people in countries where such action is deemed illegal. It is not clear where these measures are in effect.

International pressure had been building on xAI to rein in Grok after its so-called "Spicy Mode" feature allowed users to create sexualized deepfakes of women and children using simple text prompts such as "put her in a bikini" or "remove her clothes."

According to the Center for Countering Digital Hate (CCDH), Grok generated an estimated three million sexualized images of women and children in a matter of days. 

AFP