X Moves to Curb Grok’s Sexualised Images as Global Backlash Grows

X, owned by Elon Musk, says it will begin geoblocking users of its AI chatbot Grok from creating images of people in “bikinis, underwear, and similar attire”, as regulators around the world move against a surge of sexualised deepfakes generated on the platform.

The decision follows mounting criticism over Grok’s ability to produce explicit images of real people, often without their consent.

“We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis,” X’s safety team said in a statement late Wednesday.

Under the new rules, only paid subscribers will be allowed to create and edit images. Users will also be blocked from generating sexualised content “in those jurisdictions where it’s illegal”. The company did not specify which regions would be affected or how the geoblocking would function in practice.

X said it maintains “zero tolerance for any forms of child sexual exploitation, non-consensual nudity, and unwanted sexual content”.

The announcement comes as xAI’s Grok faces investigations, restrictions, and outright bans in multiple countries. In recent weeks, users exploited the chatbot’s so-called “spicy mode” to create deepfake images using prompts such as “put her in a bikini” and “remove her clothes”. The feature was widely used to generate sexualised images of women and children without consent.

On Wednesday, California Attorney General Rob Bonta announced an investigation into “whether and how xAI violated the law” by “facilitating the large-scale production” of deepfakes used to harass women and girls, including “children in nude and sexually explicit situations”.

Bonta said the probe was triggered by an “avalanche of reports”.

Regulators elsewhere are also stepping in. The United Kingdom’s media regulator Ofcom opened an investigation earlier this week, while French ministers said they had referred Grok-generated sexual content to prosecutors and to media watchdog Arcom.

Indonesia became the first country to temporarily block access to Grok, citing rising concerns over deepfake abuse. Malaysia followed with a ban of its own and said it intends to pursue legal action against X.

At the European level, the European Commission has extended a data retention order issued to X last year, requiring the company to preserve all internal documents and data related to Grok until the end of 2026.

Hours before X announced the new safety measures, Musk pushed back against the criticism, insisting he was unaware of any illegal content generated by the chatbot.

“I am not aware of any naked underage images generated by Grok. Literally zero,” Musk wrote on X. “Obviously, Grok does not spontaneously generate images, it does so only according to user requests,” he added.