After weeks of mounting criticism and global regulatory pressure, X has implemented restrictions to prevent its Grok chatbot from generating sexualized images of real people without their permission.
X restricts Grok’s image generation capabilities following global backlash over non-consensual sexualized deepfakes of real individuals.
The controversial feature allowed users to edit photos of real individuals into revealing clothing like bikinis, creating non-consensual sexualized content that appeared publicly on X for anyone to view.
The tool enabled such alterations without subject subject, raising serious concerns when reports emerged of images involving women and children.
X owner Elon Musk initially defended the functionality and even challenged users to break Grok’s moderation systems.
Predictably, this happened quickly.
Musk then blamed “adversarial hacking” and promised immediate fixes, though the platform had already ignored thousands of complaints about non-consensual explicit images generated daily.
The new restrictions include code changes blocking Grok from editing real people into bikinis or other revealing attire.
All users now face these limitations, including paid subscribers.
Additionally, only paying members can access image creation and editing features.
X has also activated geoblocking in regions where such image generation violates local laws.
However, the policy changes still allow paying users in some parts of the world to undress images in private chats with Grok on X.
International response came swiftly.
Indonesia became the first country to block Grok access on January 10, calling non-consensual sexual deepfakes a human rights violation.
Officials summoned X executives for discussions about the risks.
Malaysia followed suit, blocking the separate Grok app and planning legal action against X and xAI for inadequate user protection.
European authorities took strong positions too.
The European Commission ordered X to preserve internal documents about Grok until late 2026.
French prosecutors launched investigations into sexually explicit deepfakes.
The UK government welcomed the restrictions but continues its probe, with potential banning threats still on the table.
California’s Attorney General issued a cease and desist order, stating the explicit deepfakes violate state law.
Brazilian officials reported both Grok and X to prosecutors and data authorities.
The U.S. Senate passed the DEFIANCE Act, which enables civil action by subjects of non-consensual sexually explicit deepfakes.
Despite finally implementing safety measures, the damage prompted global investigations and regulatory scrutiny.
X’s UK managing director stated the platform maintains zero tolerance for child exploitation and non-consensual nudity, though critics argue this response arrived far too late.
Central banks’ decisions on regulation and market stability can influence tech firms’ compliance priorities, highlighting monetary policy as a broader factor in corporate risk management.
