Grok AI chatbot restricted after generating sexualized deepfakes
X announced restrictions on Grok, its AI chatbot, preventing it from editing images of real people into revealing clothing or generating explicit content. The action followed criticism over a "Spicy Mode" feature that allowed users to create sexualized deepfakes of women and children, and came hours after California's attorney general launched an investigation into xAI.
What happened
The AI chatbot Grok, integrated into X platform, featured a 'Spicy Mode' function that allowed users to create sexualized deepfake images of women and children using simple text commands such as 'dress her in a bikini' or 'remove her clothes.'
X announced on 15 January 2026 that Grok would no longer edit photographs of real people to display them in revealing clothing. The X security team stated it had implemented technological measures to prevent the chatbot from editing images of real people in revealing clothing like bikinis, and that this restriction applied to all users including paid subscribers.
Reported impact
The capability to generate sexualized deepfakes drew substantial criticism for enabling the creation of explicit content involving women and children.
Malaysia and Indonesia blocked access to the chatbot on 14 January 2026 due to the feature. The European Union also criticized the system.
Index analysis
The restriction was announced hours after California's attorney general began investigating xAI for creating 'unauthorized, sexually explicit content.'
Governance controls
X implemented technical safeguards to prevent image editing functionality that would display real people in revealing clothing.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.