Musk’s Grok under fire over ‘nudify’ image edits
The AI chatbot integrated into Elon Musk’s social platform X is being accused of enabling ‘nudify’ edits, and regulators are taking notice.
Grok, the AI chatbot built into Elon Musk’s social platform X, has been used to produce sexualised ‘edited’ images of real people, including material that appeared to involve children. In a statement cited in the report, Grok attributed some of the outputs to gaps in its safeguards that allowed images showing ‘minors in minimal clothing,’ and said changes were being made to prevent repeat incidents.
One case described a Rio de Janeiro musician, Julie Yukari, who posted a New Year’s Eve photo on X and then noticed other users tagging Grok with requests to alter her image into a bikini-style version. She said she assumed the bot would refuse, but AI-generated, near-nude edits of her image later spread on the platform.
The report suggested that the misuse was widespread and rapidly evolving. In a brief midday snapshot of public prompts, it counted more than 100 attempts in 10 minutes to get Grok to swap people’s clothing for bikinis or more revealing outfits. In dozens of cases, the tool complied wholly or partly, including instances involving people who appeared to be minors.
The episode has also drawn attention from officials outside the US. French ministers said they referred the content to prosecutors and also flagged it to the country’s media regulator, asking for an assessment under the EU’s Digital Services Act. India’s IT ministry, meanwhile, wrote to X’s local operation saying the platform had failed to stop the tool being used to generate and circulate obscene, sexually explicit material.
Specialists quoted in the report argued the backlash was predictable: ‘nudification’ tools have existed for years, but placing a powerful image editor inside a significant social network drastically lowers the effort needed to misuse it and helps harmful content spread. They said civil-society and child-safety groups had warned xAI about likely abuse, while Musk reacted online with joking posts about bikini-style AI edits, and xAI previously brushed off related coverage with the phrase ‘Legacy Media Lies.’
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!