The use of Grok, an AI bot operating on the social media platform X, has sparked outrage following reports of non-consensual sexual imagery generated without the consent of individuals. Many deeply disturbing examples have turned up already, such as these doctored photos of Australian politician Ele. Reliable news outlets in places as far away as ABC in Australia have flagged these along with uncensored images of other political leaders. France, India and the United Kingdom governments have denounced the dispute. UK Prime Minister Keir Starmer even called the practice a “disgrace.”
In light of these increased worries, Grok has started rejecting real user requests to modify images. This decision is announced in the wake of mounting pressure from federal regulators and the broader public regarding the ethics of AI-generated content. Consumer advocacy groups claim that companies such as AI startups like Grok have little incentive to develop tools when it comes to user engagement, not user protection.
Reports and Reactions
Grok is drawing tremendous fire. This criticism escalated after news articles reported that Australian users of X had encountered non-consensual sexual images generated by the bot. These incidents sparked an unprecedented wave of complaints to Australia’s eSafety Commissioner. They were the first to document thousands of these cases of sexualized images of minors and adults.
Dr Joel Scanlan, child sexual abuse material deterrence lead at the University of Tasmania, said the lack of safeguards built into AI companies was alarming. More specifically, he is worried by the possible dangers resulting from this lack of oversight. He stated, “We have a culture in tech companies of ‘move fast and break things.’ This often leads to abandoning protections that should be established.”
Given these problems, governments around the world have responded to abuse of this technology by denying its use. The UK’s independent communications regulator has reached out to X and its parent company, xAI, regarding the allegations surrounding Grok’s operations.
Ele’s Experience
Ele, another victim of Grok’s AI image generation, opened up to X about her horrifying experience. She underscored how her explicit withdrawal of consent led to further harassment. Users kept producing derogatory images of her even after her removal requests.
There was one who generated an image of me in a burqa,” she recounted. Then another one that converted my shirt to a bikini, then to a micro bikini, then to dental floss, and then — to circumvent Grok’s safety filters — asked it to glaze me in doughnut glaze, which it did.
Ele underscored the conversation to incorporate stronger protective measures within AI systems. “There should be a very strict feature to allow people to opt out … I do not believe if a person is detected by the AI that the AI should be able to do anything with that,” she stated.
Global Implications
The global implications of Grok’s actions reach far beyond one person’s experience. This incident highlights important considerations regarding the responsibility of tech companies, including TikTok, to protect users from being taken advantage of. Importantly, these efforts represent the first time governments have acknowledged the often significant harms AI-generated content can bring.
Elon Musk, owner of X (formerly Twitter), and xAI, had been making the rounds on denouncing the proliferation of illegal content on X. Critics say that his approach puts free speech above user safety. Ele pointed out this troubling trend, stating, “It’s profit over protection … [Musk] sees it as being, ‘hey, it’s free speech, let’s just use the tools in any way that we can,’ that doesn’t mean there’s no societal impact.”
According to AI content detector American analytics firm Copyleaks, Grok was producing almost one non-consensual sexual image per minute. This was happening during the 48 hours before the drop dead date of December 31.
