Musk’s xAI Tightens Grok Image Tools After Backlash From California and Europe

Elon Musk’s artificial intelligence company xAI has moved to restrict image-editing features in its Grok chatbot after a surge of sexually explicit and manipulated images triggered alarm among regulators around the world.

Late Wednesday, xAI said it had rolled out new safeguards that prevent users from editing images of real people in revealing clothing, such as bikinis. The restrictions apply to everyone, including paying subscribers, the company said in a post on X.

The move comes after Grok was used to generate hyper-realistic images of women portrayed in degrading poses, ultra-revealing outfits, or appearing bruised. In some cases, the images involved minors altered to appear dressed only in swimwear, sparking widespread criticism and calls for government intervention.

Governments from Europe to Asia have been tightening scrutiny of sexually explicit AI-generated content, and Grok has increasingly been caught in the crosshairs. Earlier this month, xAI limited Grok’s image generation and editing tools to paying users only. Even so, Reuters found that the chatbot was still able to privately produce sexualized images on request as recently as Wednesday, before the company announced its latest restrictions.

xAI also said it now blocks users from generating images of people in skimpy attire in certain locations “where it’s illegal,” though it did not specify which jurisdictions are affected.

Musk, who owns xAI and also controls X (formerly Twitter), initially appeared to brush off the controversy, responding to criticism with jokes and emojis. More recently, the company has stressed that it takes reports of child sexual abuse material seriously and enforces policies against it.

California steps in

Pressure intensified on Wednesday when California officials publicly demanded answers from xAI. Attorney General Rob Bonta said his office was seeking immediate clarification on how the company plans to stop the creation and spread of nonconsensual sexualized imagery.

Governor Gavin Newsom went further, calling on Bonta to open an investigation and hold xAI accountable.

Their statements mark the strongest response yet from U.S. officials to the growing problem of AI-generated sexual imagery circulating on X. Lawmakers and advocacy groups have also urged Apple and Google to remove Grok from their app stores.

Earlier in the day, Musk said he was “not aware of any naked underage images generated by Grok,” claiming there were “literally zero.” Neither X nor xAI directly responded to questions about the California officials’ comments. An emailed inquiry to xAI received only an automated reply: “Legacy Media Lies.”

For now, regulators appear unconvinced—and the scrutiny facing Musk’s AI ambitions is only intensifying.