After widespread backlash, X announced that Grok will no longer generate AI images of real people in sexualized or revealing clothing. The move follows investigations and suspensions in multiple countries over concerns that the tool was being used to create non-consensual and harmful images.
California Attorney General Rob Bonta said his office has opened an inquiry into whether Grok enabled the creation of illegal deepfakes, including images involving minors. Officials in Indonesia and Malaysia have already suspended the chatbot, making them the first countries to ban the service, while regulators in the UK have launched their own review.
In a statement, X said it has introduced new technical safeguards to block the editing or generation of images of real people in attire such as bikinis or underwear. The restrictions apply to all users, including paid subscribers, and additional protections are being rolled out in regions where such content is illegal.
Experts say the situation highlights growing pressure on AI companies to move faster on safety. Some argue it marks a turning point, signaling the end of looser experimentation and the start of stricter accountability for how generative tools are used.