X Removes Obscene AI-Generated Content After Government Warning
X's Grok AI tool was used to generate non-consensual obscene deepfake images of women and minors. Following government warnings from India and investigations by British regulators, X removed thousands of posts and hundreds of accounts involved in creating or sharing such content, though testing after the policy change reportedly found gaps in enforcement.
What happened
X's AI image generation tool Grok was used to create non-consensual obscene deepfake images of women and minors by applying prompts such as 'bikini' to photos of real people.
The Indian Government issued an ultimatum to X demanding removal of obscene content. Britain's communications regulator Ofcom opened an investigation, while Indonesia and Malaysia temporarily banned the Grok chatbot.
In response, X removed approximately 3,500 posts containing obscene content and deleted over 600 user accounts involved in generating or sharing such material.
Reported impact
Women and minors were targeted with non-consensual obscene deepfake images created using the Grok AI tool.
Journalists testing Grok after X announced policy changes reportedly found the tool still capable of generating obscene images, suggesting enforcement gaps.
Index analysis
Elon Musk claimed Grok only generates images when users request it and refuses illegal content, stating the tool complies with laws in each country and jurisdiction.
Tech experts noted that unlike other AI tools, Grok lacked sufficient filters against obscene content generation.
Governance controls
X announced policy changes to Grok to prevent generation of non-consensual obscene images and stated it would comply with Indian laws.
X limited image generation functionality to paid subscribers following British government objections.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.