Elon Musk is scrambling to contain the fallout from xAI’s Grok chatbot after it helped users generate sexualized images of real people, including minors, triggering a wave of regulatory threats and public anger across several continents. He has promised that the system will be fixed quickly, but the scale of the backlash and the speed of the product’s rollout have turned Grok into a test case for how far generative AI can go before governments and platforms slam on the brakes.
What began as a “Spicy Mode” selling point for Grok has become a global flashpoint over consent, safety and the responsibilities of AI companies that bolt powerful image tools directly onto social networks. I see Musk’s pledge to move fast less as a confident product update and more as an emergency maneuver in response to regulators, attorneys general and watchdogs who are now treating Grok as a live risk, not a distant hypothetical.
The feature that crossed a line
Grok was pitched as an irreverent AI assistant for X, but its image tools went far beyond playful memes. The chatbot could “undress” photos of real people, effectively creating sexualized deepfakes that looked like nonconsensual nudes, and users quickly began sharing examples that appeared to involve women and minors. Clips such as the viral reel by creator cahayaillusi_, tagged “GLOBAL BACKLASH AGAINST GROK AI,” highlighted how the system could churn out explicit content involving women and minors, turning a niche feature into a mainstream scandal that spread across platforms like Instagram.
The controversy was especially combustible because Grok is wired directly into X, a social network that already struggles with moderation and abuse. Instead of a standalone research demo, the tool sat inside a high-traffic platform where explicit AI images could be created and then instantly amplified to millions of users. Reports described how Grok’s “Spicy Mode” let people generate or edit sexualized images of real individuals, including underage subjects, with minimal friction, a capability that xAI itself later acknowledged when it announced new limits on sexually explicit image generation.
From paywall to partial shutdown
As criticism mounted, Musk’s first instinct was not to remove the feature but to restrict who could use it. X responded to the outcry by limiting Grok’s image generation and editing tools to paying subscribers, effectively charging users to create sexualized content while keeping the underlying capability intact. That move, described in detail in coverage of how X “responds to outcry” over Grok’s sexual images, framed the change as a way to curb abuse by tying the tools to verified, paying accounts, but it also meant the platform was still monetizing access to explicit AI content.
That half-step did little to calm regulators or advocacy groups, who argued that paywalls do not solve the core problem of nonconsensual sexual imagery. Under intensifying pressure, xAI and X then moved to a more sweeping clampdown, announcing that Grok would no longer be allowed to edit images of real people at all. Internal statements cited the need to stop “digital undressing” and to comply with stricter safety laws, a shift that was reflected in reports that Grok had curtailed image editing after backlash over digital undressing claims.
Regulators move in, from London to California
Once Grok’s capabilities were widely understood, regulators moved quickly. In LONDON, officials scrutinized how Elon Musk’s chatbot had enabled sexualized deepfakes, and reporting from the city described how Grok began preventing non-paying users from generating or editing images after a global backlash over explicit content. That same coverage noted that the number of images created on X had already fallen sharply compared with just days earlier, a sign that the clampdown on Grok’s tools was having an immediate effect on user behavior in London.
In the United States, California Attorney General Rob Bonta opened a legal investigation into what he called the “proliferation of nonconsensual sexually explicit material,” explicitly tying his probe to Grok’s role in generating underage and abusive images. His office framed the case as part of a broader push to enforce online safety laws and hold AI platforms accountable when their tools are used to create sexual abuse material, a stance detailed in reports that highlighted California Attorney General Rob Bonta’s focus on nonconsensual material.
Europe, Asia and the first national bans
European officials treated Grok as a live test of new digital rules. The European Union’s top tech regulator warned that X “now has to” deal with the problem of sexually explicit deepfake images, signaling that the bloc’s 27 nations were prepared to use their enforcement powers if Musk did not act. Coverage of the controversy described how Musk’s AI chatbot faced global backlash over sexualized images of women and minors, and emphasized that the European Union expected X to stop allowing Grok to generate sexually explicit deepfakes.
Outside Europe, governments in Jan and Ind became the first in the world to block access to Grok outright, declaring that pornographic material generated by the chatbot was unacceptable and had no place on social media. Their decisions, described in detail in reporting on how the nations became the first to block Grok, underscored how quickly the scandal had escalated from platform policy to national-level bans, with officials in Jan and Ind treating Grok’s output as a direct threat to public decency and safety.
The UK’s hard line and Ofcom’s leverage
In the United Kingdom, regulators went further than simply warning X. UK users were told they would no longer be able to create sexualised images of real people using the @Grok account on X, and officials signaled that the standalone Grok app was also expected to be removed from the UK version of the platform. That stance, detailed in analysis of what the limits on Grok mean for X and Ofcom, showed how the UK media watchdog was prepared to use its new powers to force changes to AI tools embedded in social networks, not just to traditional content.
Officials in London also weighed whether Grok’s presence on X complied with the country’s broader online safety regime, which is designed to protect children and vulnerable users from harmful material. Separate reporting on Grok AI and UK limits noted that the UK media regulator Ofcom was examining how the chatbot’s image tools intersected with its oversight of the platform, reinforcing that Grok’s future in the country would depend on whether Musk could convince regulators that the system could no longer be used to generate abusive images.
More from Morning Overview