✋ Grok Generated 3 Million Sexualized Deepfakes

This estimate comes from the Center for Countering Digital Hate (CCDH). In early January, X users began mass-using the xAI chatbot to undress real people in the comments. It was enough to simply ask Grok to "put her in a bikini," "remove clothes," or "replace clothes with dental floss." Additionally, the CCDH calculated that the bot created about 23,000 deepfakes involving children.

Following a wave of backlash and promises from California authorities to investigate, X restricted the generation of "adult content" and limited image creation capabilities to paid users only.

However, it was already too late. Amid the scandal, Grok was blocked in Malaysia, Indonesia, and the Philippines (the latter reversed the decision after a response from xAI). Meanwhile, regulators in the UK and Brazil are investigating the situation and are not ruling out a ban either.

@hiaimediaen